Navigation – Plan du site

AccueilNumérosXXVIIIMethods for Rock-hewn Worksite An...

Methods for Rock-hewn Worksite Analysis: The Church of Göreme n° 4b (Cappadocia): A Case Study

Anaïs Lamesa
p. 23-45

Texte intégral

I would like to extend my sincere gratitude to the Turkish Ministry of Culture and Tourism for having authorized the survey in the province of Nevşehir in 2010 (n° 07375590) as well as to the members of the French Institute of Anatolian Studies for the logistical support provided during my doctoral research, under a TÜBİTAK contract. I would also like to thank Jean-Pierre Gély, Bruno Desachy and Marie-Laure Derat for their careful proofreading, Édouard de Grave for its English proofreading and Néhémie Strupler for his support.

1Some rock-hewn structures present oddities in their layout, far from the models shared by the majority of similar monuments. These particularities challenge and question the very process of their design. While a technical survey may explain part of these singular cases, a careful comparative selection of the available methods remains paramount to draw a complete picture of the progress and organisation of these peculiar worksites.

2Göreme n° 4b is a rock-hewn church located on a ridge on the west side of the road leading to the Göreme Open Air Museum. The obvious difficulties encountered during its excavation make it an ideal case study. Its analysis sheds light on the choices made by the patron, the master builder as well as the workers’ behaviour during the digging phase. It also offers clues about the church’s occupational phases and highlights the changes from the initial project.

  • 1 We refer to the methodological article by E. Rouger (1998) dealing with the application of Harris’s (...)
  • 2 The relevance of the method applied to rock art has been debated: Russell, 2000; Russell, 2012 cont (...)
  • 3 I wish to thank B. Desachy for our discussions on the application of the software Le Stratifiant to (...)

3The combination of multiple methods provides different approaches to address the technical, economic and sociological issues of rock-hewn church worksites. A first method consists of defining sequences of the chaîne opératoire through the careful analysis of tool marks (See Bessac, 2007). Extending this first approach, an anthropological study that involves recording teams of stonemasons working traditionally will be at the core of the second method. In Cappadocia, only a few workers use non-mechanised tools. The third method experiments a technique proposed by E. Harris to determine phasing (1997). Currently this approach is used in excavations and in the archaeology of buildings.1 Sharing the intuition of a set of researchers working on African Studies, we believe that the concept may also apply to the study of structures excavated in rock.2 To carry out this analysis, we use the results obtained by the software developed by B. Desachy, Le Stratifiant.3 This computerized tool allows for the semi-automation of the reading process and produces visuals such as diagrams (Djindjian & Desachy, 1990; Desachy, 2008a; Desachy, 2008b; Desachy, 2015). At the crossroads, where these three methods meet, the decisions of the project manager and the workers come to light, and it is possible to generate a chronology for the rock-hewn worksite.

4On their own, these methods do not provide a complete picture of the church project. They are limited to the definition of digging and finishing actions, albeit as complete as possible. All aspects of the commission – site location, financing, initial choice of church plan – cannot be addressed without the gathering of a comprehensive collection of written or oral sources. This data is currently unavailable for Cappadocia. In this analysis, we focus our study on the excavation and finishing of the church’s main structural shell, excluding the realisation of the sculptured and monolithic elements.

Local Geology

5Like any worksite, the choice of raw materials is fundamental. Indeed, the rock in which the church has been excavated had a considerable impact on the worksite’s organisation and played an essential role in the choice of tools, digging strategies and in the monument’s final appearance. The church Göreme n° 4b has been excavated above what is today’s ground level in a “fairy chimney” formed by a volcanic rock, called ignimbrite. These are deposits of solidified pyroclastic flow.

  • 4 The dating of this ignimbrite layer is still being discussed: Le Pennec et al. 2005; Viereck-Goette (...)

6From the characteristics of the rock, the church appears to be clearly carved out of the so-called Zelve ignimbrite, one of the oldest ignimbrites in Cappadocia. Its colour ranges from white to yellow and it displays numerous lithic pyroclasts (inclusions), notably andesites or basalt (Le Pennec et al., 1994, pp. 66-67).4 Its emission source has been located between the cities of Nevşehir and Derinkuyu (Temel et al., 1998, p. 451).

7The volcano’s proximity - source of the Zelve ignimbrite - explains the size of the inclusions observed in the church. Pyroclastic clouds consist of rock fragments and ashes. Once emitted by the volcano, this fiery matter gradually falls down, cools and eventually hardens. The pieces of rock or pyroclasts, heavier than the ashes, settle on the ground first.

8Many pyroclasts, dark and of centimetric scale, can be seen on the church walls. Three are particularly remarkable: the first inclusion, of decimetric scale, can be seen on the nave’s north wall. The second and third inclusions, both of metric scale, are in the nave, near the vault and pew, respectively. As we will see, all three inclusions had an impact during the excavation (Fig. 1).

Fig. 1: Two inclusions of decimetric and metric scale, the Church Göreme n°4b

Fig. 1: Two inclusions of decimetric and metric scale, the Church Göreme n°4b

A. Lamesa

General Description and Dating Elements

  • 5 For descriptive convenience, the apse is considered to face east.
  • 6 This alteration was first noted by S.-A. Wallace (1991, II p. 532), who explains the irregularity o (...)

9The rock’s heterogeneous nature altered the course of the construction process and considerably modified the church’s final shape. A simple observation of the nave confirms this. The monument has a transversal nave (Fig. 2), but the shape of its vault is unusual. On the south side,5 a very regular barrel vault suddenly stops giving way to a kind of flat ceiling on the north side. In this part of the roof, delimitation with the walls of the nave is not clearly defined (Fig. 3).6 The presence of the first metric scaled inclusion explains this interruption. Similarly, a bench runs along the northern and eastern walls. This bench is of irregular aspect and widens towards the northwest of the nave. Coincidentally, we find the second metric inclusion of hard stone (Fig. 4) near this irregularity.

Fig. 2: Plan of the Church Göreme n°4b

Fig. 2: Plan of the Church Göreme n°4b

A. Lamesa and V. Dubois

Fig. 3: Details of the different elements composing the nave

Fig. 3: Details of the different elements composing the nave

A. Lamesa

Fig. 4: Bench’s irregularity to the northwest part of the nave with sledgehammer impacts

Fig. 4: Bench’s irregularity to the northwest part of the nave with sledgehammer impacts

A. Lamesa

  • 7 About these underground rooms, see Jolivet-Lévy & Lemaigre Demesnil, 2015, p. 33.

10The presence of rooms under the church could indicate that steps used to lead to a now destroyed doorway in front of the present monument (Fig. 5).7 The nave is in an awful state of conservation and has been completely gutted. Part of its western wall and the whole of its southern wall have fallen into the valley down below (Fig. 6). Three alcoves were carved into the nave. The first one is on the north wall, at man’s height. It is wide and flat-bottomed. The other two alcoves were excavated at the bottom of the eastern wall. Their bottoms are curved, and they have been identified as seats (Fig. 7) (Jolivet-Lévy, 1991, p. 90). A panel near the flat-bottomed niche on the northern wall adorns a painting depicting a holy warrior. Its execution has been dated to the 11th century (Jolivet-Lévy, 1991, p. 90).

Fig. 5: View of the Church Göreme n°4b and its supposed entrance

Fig. 5: View of the Church Göreme n°4b and its supposed entrance

A. Lamesa

Fig. 6: The apse and the southeast room of the Church Göreme n°4b’s nave

Fig. 6: The apse and the southeast room of the Church Göreme n°4b’s nave

A. Lamesa

Fig. 7: View of the apse of the Church Göreme n°4b

Fig. 7: View of the apse of the Church Göreme n°4b

A. Lamesa

11The church has a single apse on its northern part, with a flat bottom and an irregular shape, accessible by three steps. Its entrance is highlighted by a frame materialised by a relief side on the east wall of the nave. A monolithic altar flanks the eastern wall. A single alcove with a curved bottom can be found to the south of the sanctuary. A small step seems to allow one to climb to its level. This recess could therefore also be a seat. Two coats of paint on the plaster must have covered the entirety of the walls and the barrel vault of the apse. The paintings are very damaged but some of them could still be successfully identified. Of the first painted layer, a gemmed cross is the only remaining element (Jolivet-Lévy, 1991, p. 90). It is visible near the barrel vault. Surrounded by two saints that cannot be identified, the bust of Christ appears on the last layer. The face of the Christ surmounts a small alcove that was later transformed into a window. In the recessed seat, St. Basil has been identified by means of a fragment of an inscription (Jolivet-Lévy & Lemaigre Demesnil, 2015, p. 33).

12At the southern end of what remains of the nave, an irregular room, collapsed on its southern half, opens to the east. It has never been noticed before – yet it seems essential to the understanding of the church’s architectural organisation and its initial projected design (Fig. 8).

Fig. 8: View of the unfinished room in the south part of the church

Fig. 8: View of the unfinished room in the south part of the church

A. Lamesa

13Typo-chronologies of the painted ornamentations and the church’s layout help identify the phases of excavation and occupation of the monument. According to C. Jolivet-Lévy and N. Lemaigre Demesnil’s hypothesis, the digging would have been carried out shortly before the first phase of painted ornamentation – between the 7th and 9th centuries. The second layer of paintings in the apse took place between the 9th and 10th centuries (Jolivet-Lévy & Lemaigre Demesnil, 2015, p. 33). The last painted panel in the nave, seems to have been painted sometime during the 11th century.

14Other phases of occupation can be deduced. The military saint’s panel, dated to the 11th century, has been obviously destroyed by the large recess dug in the northeast corner of the nave’s northern wall (Fig. 9). The destruction of the painted panel is evidence of a late occupation phase, as the devotional object that this particular military saint used to represent got forgotten.

Fig. 9: The naves’ painting destroyed by a niche

Fig. 9: The naves’ painting destroyed by a niche

A. Lamesa

Using Tool Marks to Reveal the Chaîne Opératoire in Place

15Understanding the phase sequence of the church Göreme n° 4b worksite brings to light the master builder’s first intentions and its probable successive adaptations. A first method focuses on defining the sequencing of the church’s building through a chaîne opératoire. This theoretical model, mainly used in ethnology, aims to characterise the technical changes and the social organisation of the work (Martinelli, 1991, p. 66). Its use in our study derives from the absence of written or oral sources to describe the excavation of rock-hewn monuments in Cappadocia.

16The tasks (first degree of the chaîne opératoire), the stages (second degree) and the phases (third and last degree) describe the work sequence. Tool marks on the wall constitute the princeps element by which the model can be constructed. They emanate from a tool that is used for a specific task. They form a tool cut, a series of tool marks, which represents the action the tool performed on the wall to complete that task. They can be broken down into two elements: the longitudinal mark and the final mark. The longitudinal mark (i.e. groove) corresponds to the tool mark left by the tool when it is in motion. The final mark (i.e. impact) is made by the tool’s cutting edge.

17The chaîne opératoire’s first sequence can be deduced by identifying the marks of the three main tools used during the excavation:

  • 8 We have distinguished the picks according to the length of their handle because in Cappadocia, two (...)

18The predominant tool marks on the church’s wall originate from a tool with a short handle (Fig. 10).8 The grooves they typically leave behind are narrow, short, and evenly spaced. They form a distinctive radial and slanted cut, a feature of the short-handled pick. This cutting process is used to straighten up walls. The series of the longitudinal impacts forms an arc. The stonecutter operates in profile, and the marks result from the movement of his tool. This cutting is carried out voluntarily for functional or aesthetic purposes.

Fig. 10: A long-handled pick and a short-handled pick held by stonemason K. Manap

Fig. 10: A long-handled pick and a short-handled pick held by stonemason K. Manap

A. Lamesa

  • 9 Bessac, 1986, pp. 14-24 for the long-handled pick or stonecutter pick.
  • 10 On radial, slanted, and linear cutting: Bessac, 1986, pp. 19-20.

19A second type of marks also result from longitudinal impacts of a throwned tool, a long-handled pick9 in this case. These differ from the short-handled pick tool marks with distinctive wide and long grooves and cuttings of various types – radial, slanted, linear – may be observed throughout the monument. For the radial type, the quarryman handles his tool parallel to the wall. The longitudinal impact series forms an arc. The quarryman operates with the bed rock he aims to remove in front of him. This cutting is often associated with linear cutting. These are the final impressions left on the rock mass by the worker. This cutting leaves round or square impacts (depending on the tool’s cutting-edge wear) in vertical line patterns. Both types of cuttings can be classified as destructive.10 In the church Göreme n° 4b, tool marks of long-handled picks can be seen in the south-eastern hall and at the bottom of the nave’s western wall (Fig. 11).

Fig. 11: Long-handled picks’ impacts, west wall of the nave

Fig. 11: Long-handled picks’ impacts, west wall of the nave

A. Lamesa

20The third type of tool marks also results from a thrown percussion tool: an axe with two cutting edges (one vertical and the other horizontal) named polka or a carving axe (Bessac, 1986, pp. 39-52, 53-59). It is not easy to distinguish these two tools by their marks. In the church Göreme no. 4b, some of these tool marks are clearly visible on the relief side of the apse frame on the nave’s eastern wall (Fig. 12). The tool marks are distributed horizontally, trademarks of a polka. The chosen cutting is of the slanted type. This tool was exclusively used on the eastern parts of the nave, the cornice running along the nave, the intrados and mouldings of the triumphal arch, and the upper part (covered with paintings) of the apse (Fig. 13) to finish the flattening of the walls.

Fig. 12: Polkas’ tool marks on the relief side on the east wall of the nave

Fig. 12: Polkas’ tool marks on the relief side on the east wall of the nave

A. Lamesa

Fig. 13: View of the east wall of the nave: junction between the short-handled picks’ tool marks and the polkas’ tool marks

Fig. 13: View of the east wall of the nave: junction between the short-handled picks’ tool marks and the polkas’ tool marks

A. Lamesa

21The layers’ superimposition makes a hewing chronology possible: the tool marks of the long-handled pick are covered by the impacts of the short-handled pick (Fig. 14). The tool marks of the polka very clearly overlap those of the short-handled pick.

Fig. 14: West wall of the nave of the Church Göreme n°4b, radial and slanted cutting by short-handled pick

Fig. 14: West wall of the nave of the Church Göreme n°4b, radial and slanted cutting by short-handled pick

A. Lamesa

22This sequence of tasks, stages and phases may be translated as follows (Table 1):

Table 1

Phases

Stages

Tasks

Actions

Tools

Location

Realising the project

Roughing out the structural work

Roughing out the wall

Destructive radial and slanted cutting

Long-handled pick

Western wall of the nave and eastern room

Finishing the project

Finishing the structural work

Straightening the wall

Radial and slanted cutting

Short-handled pick

West and north walls of the nave

Flattening the wall

Skewed slanted cutting

Polka

Eastern wall of the nave

23The observation of hewing accidents further details the chaîne opératoire’s first sequence. In the case of the church Göreme n° 4b, the workers carved an ignimbrite with many hard inclusions. Some of them could be removed when the wall was straightened with a short-handled pick. Others had to be levelled, perhaps using a carving axe or a polka. Two, however, could not be removed and show tool marks on their surface, which allows us to further define the stone workers toolbox.

24On the inclusion of the north wall, which is of decimetric scale, two types of marks are clearly visible. The first are around 0.5 to 0.7 cm in diameter. These marks come from the cutting edge of the long-handled pick. The second type does not correspond to a tool that has already been observed: they are composed of coarse lines caused by the bursting of the rock (Fig. 15). These tool marks result from the use of a sledgehammer to remove the inclusion. The use of this tool is explained by the very nature of the inclusion – a hard stone (basalt?) and the inefficiency of the traditional tools used for carving – in this case the long-handled pick.

Fig. 15: Inclusion of decimetric scale where the long-handled pick and sledgehammers’ tool mark appear

Fig. 15: Inclusion of decimetric scale where the long-handled pick and sledgehammers’ tool mark appear

A. Lamesa

25On the inclusions of metric scale, visible both on the ceiling and floor of the church, the tool marks are less clear, but the shape of the longitudinal impacts seems to confirm the bursting by a sledgehammer. The lines are coarse and irregular, and the angles are sharp.

  • 11 Additions to the table are indicated by the colour grey.

26The chaîne opératoire’s second sequence therefore concerns the treatment of inclusions, which the workers had to take into consideration during the carving process. This sequence completes the relative chronology (Table 2).11

Table 2

Phases

Stages

Tasks

Actions

Tools

Location

Realising the project

Roughing out the structural work

Roughing out the wall

Destructive radial and slanted cutting

Long-handled pick

Western wall of the nave and eastern room

Removing hard inclusions

Burst cutting

Sledgehammer

Inclusions of decimetric and metric scale

Managing the two hard inclusions

Abandonment of the destruction attempt

North wall of the nave and vault

Finishing the project

Finishing the structural work

Straightening the wall

Radial and slanted cutting

Short-handled pick

West and north walls of the nave

Removing hard inclusions

Destructive cutting

Short-handled pick

Inclusions of

centimetric scale

Flattening hard inclusions

Polka

Inclusions of

centimetric scale

Flattening the wall

Skewed slanted cutting

Polka

Eastern wall of the nave

27This second sequence also highlights the difficulties encountered by the workers during the excavation process and the technical solutions implemented to get around them. It illustrates the choice of the master builder who modified the final appearance of the work by leaving three hard inclusions in place.

28A first limitation to sequencing a chaîne opératoire is the impossibility to establish what types of measuring or drawing tools were used by the workers on the site. However, the vault’s perfect cradle shape seems to indicate the use of a tool that is able to carry out perfect semicircles.

29The second limit derives from specific tasks in the chaîne opératoire, the purpose of which is to precisely make tool marks disappear. An example is the fine smoothing of the walls: this technique is commonly used in Cappadocia, where the rough/smooth contrast can be played with (Fig. 16) (Lemaigre, 2010, pp. 69-72; Jolivet-Lévy 2007, pp. 52-60; Jolivet-Lévy & Lemaigre Demesnil, 2015, p. 118). This fine finishing of the walls was applied around the nave’s alcoves and passages (see Fig. 13).

Fig. 16: Ceiling of the Church of St. Basil/Uçhisar n°3. Short-handled pick’s tool marks highlight the smoothing of the cross

Fig. 16: Ceiling of the Church of St. Basil/Uçhisar n°3. Short-handled pick’s tool marks highlight the smoothing of the cross

A. Lamesa

30Thanks to the very accurate observation of the protected spaces (e.g. under a moulding), it is possible to detect the use of a rasp or of an abrasive (Bessac, 1986, pp. 203, 264). It is difficult to distinguish the two tools as they both leave fine grooves on the wall (Bessac, 1986, p. 207). In Cappadocia, we were able to observe some of these marks on a moulding of a monumental tomb in the Davutlu Bucağı Valley (Fig. 17) (Lamesa, 2016, p. 416). However, these finishing techniques do not seem to have been used in the case of the church Göreme n° 4b. The space to be smoothed is far too large. Abrasives or rasps are typically used for mouldings and no streaks could be found.

Fig. 17: Using a rasp or an abrasive to smooth out the moulding of the T05 DB tomb located in the Davutlu Valley Bucağı at Mazıköy

Fig. 17: Using a rasp or an abrasive to smooth out the moulding of the T05 DB tomb located in the Davutlu Valley Bucağı at Mazıköy

A. Lamesa

Ethnoarchaeology in Cappadocia: Possible Tracks to Explore

  • 12 It should be noted that stone workers most often use a pneumatic hammer for hewing and cutting the (...)

31In order to identify the motions associated with these finishing stages, we have chosen to call upon ethnoarchaeology, through an experimental approach and the observation of practices (Gallay, 2011, pp. 298-299). Alain Gallay stressed the need to set up a quantitative study (2011, p. 301). This is difficult to carry out in Cappadocia since rock art is no longer practised. However, a couple of stonemasons still hew conservation deposits or rock rooms (rooms, refectories) for luxury hotels using traditional tools at certain stages (Öztürk, 2009; Lamesa, 2011; Öztürk, 2017; Lamesa, 2018).12 We therefore focused on a qualitative study using direct observation and interview techniques. This is a case study and the collected data must be considered as possible solutions.

  • 13 J.-Cl. Bessac mentions the replacement of the squeegee with a very sharp chisel.

32We followed the stonemasons’ team led by Kemal Manap. For a smooth rendering of a block used in a wall, the master/usta uses the blade of the straight edge as a scraper to remove the roughness left by the tool’s impact (Fig. 18). He then finishes smoothing with an artificial abrasive (sandpaper). This straight-edged scraped cut can also be performed using a wide chisel with a sharp, straight edge (Bessac, 1986, p. 190, see also Lamesa, 2011, p. 185).13 The resultant effect matches what could be observed on the church walls. The tool marks of the straight edge are still visible depending on the incidence of light, but from the front, the wall appears smooth.

Fig. 18: K. Manap using the polka as a scraper

Fig. 18: K. Manap using the polka as a scraper

A. Lamesa

  • 14 Additions to the table are indicated by the colour grey.

33We therefore propose to insert this solution (Table 3) into the hewing process of church Göreme n° 4b, but we should stress that for now this remains a hypothesis14:

Table 3

Phases

Stages

Tasks

Actions

Tools

Localisation

Realising the project

Roughing out the structural work

Roughing out the wall

Destructive radial and slanted cutting

Long-handled pick

Western wall of the nave and eastern room

Removing hard inclusions

Burst cutting

Sledgehammer

Inclusions of decimetric and metric scale

Managing the two hard inclusions

Abandonment of the destruction attempt

North wall of the nave and vault

Finishing the project

Finishing the structural work

Straightening the wall

Radial and slanted cutting

Short-handled pick

West and north walls of the nave

Removing hard inclusions

Destructive cutting

Short-handled pick

Inclusions of

centimetric scale

Flattening hard inclusions

Polka

Inclusions of

centimetric scale

Flattening the wall

Skewed slanted cutting

Polka

Eastern wall of the nave

Scraped cutting

Polka or chisel

Eastern wall of the nave

34Three other pieces of information obtained through ethnoarchaeology will later be useful for our demonstration.

35Thanks to the photographs published by F. Öztürk (2009, res.17-18) or collected during interviews, we notice that (1) the rubble allowed the stone worker to climb to ceiling level. In the picture (Fig. 19), the tool marks (2) are distributed horizontally at the level of the upper part of the wall (slanted cutting) and follow an arc at the centre of the wall (radial cutting). The extremities of the room (ceiling, corners and floor) require a precise and meticulous cutting to be carried out, which is impossible with slanted and radial cutting. It appears that (3) the stonecutter’s work was carried out in sections, as rock reserves were gradually being removed (Öztürk, 2009, including res.17).

Fig. 19: 1980s photograph of Mustafa Işık digging the depot near the Göreme Museum

Fig. 19: 1980s photograph of Mustafa Işık digging the depot near the Göreme Museum

Anonymous, given to A. Lamesa in 2008

36Data collection through interviews and experimentation is still in its infancy in Cappadocia. Hence the need to combine this information with other archaeological methods to confirm or disprove it. The hypotheses formulated thanks to ethnoarchaeology can be discussed in a relevant way only by cross-matching with other methods.

The Phasing of a Wall: A Complex but Incomplete Vision of the Work

  • 15 The team working at Lalibela applies this technique of reverse sequencing by studying the superimpo (...)
  • 16 Wall measurements: 172 cm long, 214 cm high (measurement taken on the step) and 246 cm high (at the (...)

37A stratigraphic reading of the tool marks left on the walls further refines the church’s building sequence and confirms archaeologically the data collected in ethnoarchaeology. This third method, based on the principles advocated by E. Harris, needs to be adapted to the characteristics of the rock-hewn space. Unlike stratigraphic units, layers accumulate one on top of the other according to the Law of Superposition (the oldest layers being covered by the latest). The newest tool marks destroy the oldest ones by physically removing a layer from the wall. We therefore propose to call these layers Negative Units instead of Stratigraphic Unit (Fauvelle-Aymar et al., 2010, p. 1138).15 This is a particularly time-consuming method to deploy over the entire rock-hewn monument. It should therefore only be used in very specific and carefully selected promising areas. In our case, we chose the western wall of the nave, of which only the northern part remains.16 This wall is essential insofar as it has not received the same finishing touches as its eastern counterpart. Moreover, it is located at the junction of the two ceilings, separated by the metric scale inclusion of hard rock and at the level of the metric inclusion on the floor.

Fig. 20: Delimitation of Negative Units

Fig. 20: Delimitation of Negative Units

A. Lamesa

38To properly define the phasing of the negative strata (Table 4), it is necessary to delimit and correctly describe the test wall by identifying the type of tool marks, their orientation and their spacing (Fig. 20):

Table 4

Name

Series of marks orientation

Spacing

Stone worker’s position

Tool type

Cutting

1

South to north,

descendant

Less than 1 cm

Parallel to the wall,

northerly

Short-handled pick

Radial and slanted cutting

2

North to south,

descendant

Less than 1 cm

Parallel to the wall,

southerly

Short-handled pick

Radial and slanted cutting

3

Almost vertical,

top-down

Less than 1 cm

Parallel to the wall, southerly

Short-handled pick

Linear cutting

4

South to north

descendant

Less than 1 cm

Parallel to the wall,

northerly

Short-handled pick

Radial and slanted cutting

5

South to north,

descendant

Less than 1 cm

Parallel to the wall,

northerly

Short-handled pick

Radial and slanted cutting

6

South to north

descendant

Less than 1 cm

Parallel to the wall,

northerly

Short-handled pick

Radial and slanted cutting

7

Anomaly

8

North to south

descendant

Between 1 and 2 cm

Parallel to the wall,

southerly

Long-handled pick

Radial and slanted cutting

9

South to north

descendant

Less than 1 cm

Parallel to the wall,

northerly

Short-handled pick

Radial and slanted cutting

10

South to north

descendant

Between 1 and 2 cm

Parallel to the wall,

northerly

Long-handled pick

Radial and slanted cutting

11

South to north

descendant

Between 1 and 1,5 cm

Parallel to the wall,

northerly

Short-handled pick

Radial and slanted cutting

12

South to north

descendant

Between 1 and 2 cm

Parallel to the wall,

northerly

Long-handled pick

Radial and slanted cutting

39To this stratification, we add a few scattered or grouped tool marks that do not actually perform series (Table 5).

Table 5

Name

Series of marks orientation

Location

Stone worker’s position

Tool type

A

South to north

descendant

Negative Unit (NU) 1

Parallel to the wall,

northerly

Short-handled pick

B

North to south

descendant

NU 9

Parallel to the wall,

southerly

Short-handled pick

C

North to south

descendant

NUs 3, 7, 8

Parallel to the wall,

southerly

Short-handled pick

D

North to south

descendant

NU 8

Parallel to the wall,

southerly

Short-handled pick

E

North to south

descendant

NU 9

Parallel to the wall,

southerly

Polka

40By mechanically applying the Principle of Stratification of Negative Units, an automation made possible by the software Le Stratifiant, a striking pattern emerges. While nine phases appear on the Harris matrix, Negative Units 1 and 8 do not seem to correspond to the same period. A lost chronological relationship is therefore revealed between these two units (Fig. 21).

Fig. 21: Harris diagram of Negative Units taken from Le Stratifiant software, non-phased.

Fig. 21: Harris diagram of Negative Units taken from Le Stratifiant software, non-phased.

41Can these results be cross-matched with the sequencing of chaîne opératoire data and the information obtained by ethnoarchaeology?

42Units 1 and 9 form two separate floors. Both cases are related to the straightening of the wall after excavation, as shown by the extremely regular cutting made by a short-handled pick. The two layers’ overlapping in the north is explained by the proximity of the north wall. The worker wanted to accurately shape the junction between the north and west walls. He therefore reworked this critical and delicate area of the wall. The floor-by-floor breakdown shows that a clearance operation was carried out between the two floors. These two units represent two chronologically distinct tasks.

43After completing Unit 1, the stone worker seems to have taken over the centre of the wall. This is visible at the level of Units 4 and 5, which appear to us to be synchronous. The greater depth of tool marks at Negative Unit 6’s level and the change in the marks’ angle prove that this is a distinct third level in the phase. This last unit seems to be the leftover of a task, which was then subsequently covered by Unit 9. Could it be the presence of a first floor formed by the rubble? The presence of an important inclusion? In any case, the wall was cut again, and this action required the worker to turn around and slightly change position.

44Several indications tend to prove that Negative Units 2, 3 and 7 were carried out between two tasks of straightening the wall, represented mainly by Units 1 and 9. The mason did not intervene at ceiling level, demonstrating that he had completed the ceiling’s shaping. The stone worker’s position, inferred from the radial and slanted cuttings, proves that he was slightly elevated from the current floor. The clearance of the space had therefore not been fully completed. It is difficult to explain the reasons that led the stonemason to take up the wall in the opposite direction: the need to straighten the wall? The levelling of imperfections due to inclusions? The linear cut at Unit 3 could indicate the presence of an obstacle to the south. Could it be the hard inclusion in the ceiling that prevented the worker from easily manipulating his tool? The distorted Unit 7 indicates in any case that the worker had encountered some kind of difficulties in cutting at this level of the wall.

45Unit 11’s tool marks do not share the same spacing as those of Negative Unit 9. However, the same tool has been used, indicating that the cutting had been resumed consecutively to the previous task. The lengthening of the worker’s movement and the significant penetration of the tool into the wall at the level of Negative Unit 11 further demonstrate that the worker was performing his action close to ground level. The change of tool, a long-handled pick, at the level of Negative Unit 12 confirms this hypothesis. After this first ground level, the worker used a long-handled pick to shape a step in the immediate vicinity of the hard inclusion of metric proportion. This allowed him to make the bench appear. It is certain that at this point the space had been completely cleared out.

46Finally, Negative Unit 10 also shows signs of resuming work. The change of tool, a long-handled pick, could indicate the presence of a rock reserve close to the north wall. The need to protect the wall junctions as much as possible in order to avoid any irreversible work-related accident explains this strategic choice. The worker therefore favoured cutting the lower northern corner of the wall in a second step, once the entire area had been cleared.

47In the case of the Units A, B, C, D, this could be a one-time intervention to remove inclusions. This hypothesis is supported by the presence of inclusions of centimetre scale near A, B and C. For D, there is no guarantee, but the use of the short-handled pick seems to be associated with this task. Moreover, the traces are oriented in the same way as C, contrary to those of the Negative Unit 8. We believe that the work of levelling Unit D has been carried out in the same continuity as Unit C. The interpretation of Unit E is more delicate, the use of the polka in the centre of the wall could also indicate a flattening of inclusions.

48The relationship between Negative Unit 1 and Negative Unit 8 remains to be explained. The latter predates Units 7, 9 (and therefore Negative Unit 11). The use of a long-handled pick could be the key to this mystery, the unit would be a reminiscence of a first phase of excavation, prior to all the other units visible on the wall. The fact that the worker is facing south to carry out this digging seems to indicate that a space had already been created.

49One could thus define a chaîne opératoire thanks to this stratification of tool marks. It is however not possible to place Unit 11 and Unit 10 in a chronological stratigraphic sequence since the worker could first work on the task of Unit 10 or on the task of Unit 11. Arbitrarily, and for convenience, we will place task 11 before task 10 in Table 6 below.

Table 6

Phases

Stages

Tasks

Actions

Tools

Realising the project

Roughing out the structural work

Roughing out the wall, Unit 8

Destructive radial and slanted cutting

Long-handled pick

Finishing the project

Finishing the structural work

Straightening the wall, Unit 1

Radial and slanted cutting

Short-handled pick

Removing hard inclusions, Unit A

Destructive cutting

Short-handled pick

Straightening the wall, Units 4 et 5

Radial and slanted cutting

Short-handled pick

Straightening the wall at the ground floor, Unit 6

Radial and slanted cutting

Short-handled pick

Removing the rubble

Removing the rubble

Basket, shovel

Straightening the wall, Unit 2

Radial and slanted cutting

Short-handled pick

Straightening the wall, Unit 3

Linear cutting

Short-handled pick

Distorted area, unreadable reading, Unit 7

Removing hard inclusions, Unit C et D

Short-handled pick

Removing the rubble

Removing the rubble

Basket, shovel

Straightening the wall, Unit 9

Radial and slanted cutting

Short-handled pick

Removing hard inclusions, Unit E

Polka

Removing hard inclusions, Unit B

Straightening the wall at the ground floor, Unit 11

Radial and slanted cutting

Short-handled pick

Realising the project

Roughing out the structural work

Roughing out the wall, Unit 10

Destructive radial and slanted cutting

Long-handled pick

Roughing out the wall, Unit 10

Destructive radial and slanted cutting

Long-handled pick

50Thanks to the phasing in the software Le Stratifiant, a new Harris diagram organized according to the previously mapped chronology (Fig. 22) may be proposed.

Fig. 22: Harris diagram of the Negative Units taken from the software Le Stratifiant, phased

Fig. 22: Harris diagram of the Negative Units taken from the software Le Stratifiant, phased

51Phasing by stratigraphy reading has many advantages. The main one being the systematic record of all series of marks on a given wall. Recording the observations in Le Stratifiant then makes it possible to verify the reading of these Negative Units by pointing out redundancies and logic problems. With quick access to a graphical representation of the data, the software helps the archaeologist’s reasoning and streamlines the formulation of new hypotheses which could potentially lead current field work to be redirected on the spot.

52In the framework of a study specific to rock-hewn monuments, the case of Negative Unit 8 is particularly interesting. It is commonly accepted that the series of marks on the wall are evidence of the masons’ work from top to bottom. The archaeologist therefore tends to consider the high series of marks as the oldest. In this case, it is not Unit 1 but Unit 8, in the centre of the wall, which is the oldest. This fact gives a different orientation to the view of the space under survey. The transcription by means of graphs reveals the divergent hypotheses and opens up a new field of questioning: why did the worker not take over the wall completely and leave part of it in its most unfinished phase? Does this mean that the area further to the south, now destroyed, was waiting for a finishing phase when the church was delivered to the commissioner? Did the presence of the hard inclusion in the ceiling, still represent, during the finishing of this wall, an obstacle to the good progress of the worksite?

53For the church Göreme n° 4b, applying a stratigraphic reading at wall scale illustrates the masons’ working process. The limits of the different levels of ground reached during excavation, the organisation of the site (the stonecutter has to stop clearing) and the work sequences (the tools used are visible on the graph) are thus highlighted. In the same vein, this very accurate reading of the wall reveals the difficulties that a worker encountered and the solutions he implemented to solve them while shaping the space.

54However, this automated stratigraphic work remains incomplete. Without the ethnoarchaeological study, it would have been impossible to understand the clearing tasks during excavation, which are essential to draw a full chaîne opératoire. Similarly, the focus on a given space obscures the plurality of tools used by the stonemason and the singularity of each worker. Phasing by stratigraphic reading cannot therefore do without other methods such as tracer analysis and ethnoarchaeology. This method allows the setting of limits (in a deductive study) which by crossing information obtained by other approaches becomes an inductive study.

The Worksite of Church Göreme n° 4b: A Modified Initial Project?

55We now wish to take a broader look at the final architectural aspect of the monument, which is rare. Articulating this aspect with the results obtained previously argues in favour of a final project amended during the building process (see Fig. 2) (Jolivet-Lévy & Lemaigre Demesnil, 2015, p. 33).

56A simple observation of the monument reveals an uncommon layout and suggests a first modification of the project. The unnatural transition between a particularly well-made transverse vault and the irregular shape of the ceiling to the north of the nave is obvious. Thanks to the technical study, it could be proved that the hard inclusions of decimetric and metric scale were subject to attempts of destruction by the stone workers. The efforts made and the result shows that a decision, taken by the master builder, stopped the work and initiated a reflection on the integration of these hard inclusions in the final project.

57As R. Ousterhout notices, the plan of the apse is peculiar (2017, p. 430). It is rectangular on the north side and curved on the south side. This irregularity links to a second modification, decided by the master mason who favoured avoiding the hard inclusion altogether by considering its oblique disposition into the rock. He therefore had a straight wall built on the north side, a precaution that was superfluous over on the south side.

58The unfinished rooms’ presence to the church’s southeast finally pleads in favour of a third modification decided upon during the construction. We believe that this room is a part of the church project. The apse’s frame has been perfectly executed and the first task in the finishing stage of the work (slanted cutting with a polka) overflows back (see Fig. 12). This shows that the protruding framing is anterior to the finishing of the church. The entrance of the unfinished room is made during hewing. In addition, its entrance is marked by a pediment surmounted by a rectangular corbel and is inscribed in an opening in the shape of an arc (Fig. 23). The corner of the door can also be seen. Paradoxically, we observe a destructive radial and slanted cutting and a linear cutting characteristic of the use of a long-handled pick on its interior walls. This tool and the cuttings associated with it were identified during the construction phase of the work. The workman stopped while he was obviously shaping the room. The entrance had already been regularised and was in the process of being finished. This can be seen by the levelling of the inclusion at the level of the pediment and by the partial destruction of the nearby inclusion (Fig. 23). Similarly, the mouldings surrounding the door are well done. It is impossible to define what could have been the role of this southeast room: a second apse? A parekklesion (annex chapel)?

Fig. 23: Inclusions close to the southeast room’s entrance

Fig. 23: Inclusions close to the southeast room’s entrance

A. Lamesa

59The choice to focus on the northern part of the church seems surprising at first glance; it would, in fact, have been more logical not to continue digging beyond the hard-metric scaled inclusion of the ceiling and to make do with the spaces already available, namely the southern part of the church. However, geological problems also prevented the workmen from properly shaping the southern space, which would explain the master builder’s plan to have the two metric scaled inclusions removed in order to obtain a finalised part in the north.

60Thanks to all the data collected, the sequence of the workers’ actions and the modifications decided by the master builder can be chronologically sequenced. (1) The workers first completed the southern part of the nave, a process that was part of the initial project. As they moved northwards, they encountered several metric scale inclusions that prevented them from progressing. (2) While the shaping of the southern part is well advanced but before the finishing of the naves’ walls, a workman begins the realisation of the south-eastern room. Unfortunately, the quality of the rock stops him and this room is left unfinished. (3) The decision is made to amend the original plan of the church, i.e. a transverse nave covered with a barrel vault. The team then concentrates on the northern part of the nave. (4) After having succeeded getting through the metric inclusions to the northwest of the nave, an operation that must have required at least a two-person team, (5) a space is created, and hewing is resumed. (6) Then a worker heads east, and a second worker begins the phase of making the space to the west; the quantity of rubble had to be sufficient and the space between the inclusion of the ceiling and the north wall of the nave sufficient to start the ceiling. (7) The worker digging eastwards then finds the inclusion again. (8) The master builder then makes the decision to change the shape of the apse to avoid the hard vein, which explains the unusual shape of the apse. He chooses to carry out the enhancement of the naves’ north-eastern part. (9) At the same time, the protruding framing in the middle of the nave is committed in order to visually identify the northern part of the church as the sacred part of the monument. (10) The remaining of the naves’ eastern wall is taken over before the finishing and sculpting phases are carried out. The location of the finishing tasks, observed on the north and northeast walls of the nave and vault, confirms the importance attributed by the master builder to the northern part. This additional indication tends to prove that at the end of the worksite, the southern part of the church had lost all its appeal.

Conclusion

61The detailed study of the church Göreme n° 4b has shown that these rock-hewn structures retain the memory of the changes they have undergone over time. Thanks to the rocky nature of the monument, these changes can be included in the succession of the worksite phases, more precisely than those observed in a built monument. They highlight the decisions made by the master builder when problems were encountered during the shaping of the church, as well as the organisation of the team and the distribution of their work in the hewn space (Fauvelle-Aymar et al., 2010, pp. 1138-1139).

  • 17 On the question of unfinished works and sculptures, see the recent reflections of Vidya Dehejia & P (...)

62A part of the church n° 4b was left unfinished: this is not exceptional, other churches such as the church n° 28 of Göreme (northern apse) or the church of Bahçeli (western part of the northern nave) have the same characteristic (Lamesa, 2016, plates 5, 15). Delivering a monument to the patron does not imply that all the finishing touches have to be carried out or that certain areas have to be completed.17 The choice to proceed with the painted ornamentation while the monument was yet to be completed suggests that this “abandonment” of the excavation was not decided by the master builder but rather by the patron himself. In the case of church no. 4b, was the final choice to carry out the delicate finishing touches on the north-eastern part of the church not made by the patron? Saving in human resources for a saving in costs  – the initial contract having most certainly been subject to “amendments”  –  the patron also influenced the project manager in these decisions. In any case, this explains the dichotomy between the northern and southern parts.

63The methods proposed in this paper and the resulting reflections could be extended to other monuments. The traceology approach also needs to be cross-referenced with other methods that are more anthropologically centric. In Cappadocia, there is a need to collect more testimonies of stonemasons, to carry out experimental archaeology on ignimbrite and to observe the construction of supply sites and rock-hewn hotels. Although partly lost, the knowledge and skills related to traditional stone-cutting techniques need to be collected.

  • 18 On a first example of comparative work, see Lamesa & Gervers (forthcoming).

64Finally, other areas, such as Ethiopia, where rock church sites are still being hewn, require special attention. The establishment of comparative analytical grids will, I hope, in the future, provide new elements that will most certainly broaden the scope of our reflection.18

Haut de page

Bibliographie

Bessac, J.-Cl. (1986). L’outillage traditionnel du tailleur de pierre de l’Antiquité à nos jours. Paris: Edition du Centre national de la recherche scientifique.

Bessac, J.-Cl. (2007). Le travail de la pierre à Pétra: technique et économie de la taille rupestre. Paris: Édition Recherche sur les Civilisations.

Dehejia, V., & Rockwell, P. (2016). The Unfinished: Stone Carvers at Work on the Indian Subcontinent. New Delhi: Roli Book.

Desachy, Br. (2008a). De la formalisation du traitement des données stratigraphiques en archéologie de terrain, (1). Aperçu historiographique – note méthodologique (Doctoral thesis). Université Paris Panthéon-Sorbonne, Paris.

Desachy, Br. (2008b). Le Stratifiant, un outil de traitement des données stratigraphiques. Archeologia e calcolatori, 19, 187-194.

Desachy, Br. (2015, January 29). Le Stratifiant – téléchargement et documentation. Retrieved from https://abp.hypotheses.org/3965

Djindjian, Fr., & Desachy, Br. (1990). Sur l’aide au traitement des données stratigraphiques des sites archéologiques. Histoire & Mesure, 5(1-2), 51-88.

Fauvelle-Aymar, Fr.-X., Bruxelles, L., Mensan, R., Bosc-Tiessé, Cl., Derat, M.-L., & Fritsch, E. (2010). Rock-cut Stratigraphy: Sequencing the Lalibela Churches. Antiquity, 84(326), 1135-1150.

Gallay, A. (2011). Pour une ethnoarchéologie théorique. Paris: Editions Errance.

Harris, E. C. (19972). Principles of archaeological stratigraphy. London-San Diego: Academic Press.

Jolivet-Lévy, C. (1991). Les églises rupestres byzantines de Cappadoce. Le programme iconographique de l’abside et de ses abords. Paris: Éditions du Centre national de la recherche scientifique.

Jolivet-Lévy, C. (2007). La Cappadoce aux VIIe-IXe siècles: quelques nouveaux témoignages archéologiques. In A. C. Quintavalle (Ed.), Medioevo Mediterraneo: l’Occidente, Bisanzio e l’Islam. Atti del Convegno Internazionale di Studi, Parma, 21-25 settembre 2004 (pp. 52-60). Milan: Electra.

Jolivet-Lévy, C., & Lemaigre-Demesnil, N. (2015). La Cappadoce. Un siècle après G. de Jerphanion. Paris: Geuthner.

Lamesa A. (2011). Détermination des intervenants de chantiers d’églises rupestres en Cappadoce médiévale, VIe-XIIIe siècle: méthode d’analyse. In J. Lorenz, & J.-P. Gély (Eds.), Carrières et construction. vol. V - Actes du colloque Carriers et bâtisseurs dans la période ante-industrielle en Europe et les régions limitrophes, organisé par le C.T.H.S. à Bordeaux du 20 au 25 avril 2009 (pp. 177-190). Paris: Éditions du CTHS.

Lamesa A. (2016). D’une Cappadoce à l’autre (IIIe siècle av. J.-C. – Xe siècle ap. J.-C.: questions historiques, archéologiques et géographiques (Doctoral thesis). Sorbonne Université, Paris.

Lamesa A. (2018). Rock-cut tombs and churches in Cappadocia during the Roman and Byzantine Periods. The Analysis of Chaîne Opératoire to understand the Economy and Sociology of the Building Sites. In M. Godon (Ed.), Merging Techniques and Cultures. Technical Approaches in Archaeology, Istanbul 24-25 November 2015 (pp. 75-83). Istanbul: Institut Français d’Études Anatoliennes-George Dumézil, Diffusion De Boccard.

Lamesa A., & Gervers, M. (forthcoming). How to define the “Status” of stonemasons employed in a rock-hewn structure worksite in the Medieval Times? Reflexions and hypotheses over the long-time span. In Ch. Courault, D. Maschek, J. A. Domingo, & S. J. Barker (Eds.), From Concept to Monument: Time and Cost of Construction in the Ancient World. Oxford: Archeopress.

Lemaigre Demesnil, N. (2010). Architecture rupestre et décor sculpté en Cappadoce (Ve -IXe siècle). Oxford: Archeopress.

Le Pennec, J.-L., Bourdier, J.-L., Froger, J.-L., Temel, A., Camus, G., & Gourgaud, A., (1994). Neogene Ignimbrites of the Nevsehir Plateau (Central Turkey). Stratigraphy, Distribution and Source Constraints. Journal of Volcanology and Geothermal Research, 63, 59-87.

Le Pennec, J.-L., Temel, A., Forger, J.-L., Sen, S., Gourgaud, A., & Bourdier, J.-L. (2005). Stratigraphy and Age of the Cappadocia Ignimbrites, Turkey: Reconciling Field Constraints with Paleontologic, Radiochronologic, Geochemical and Paleomagnetic Data. Journal of Volcanology and Geothermal Research, 141, 45-64.

Martinelli, Br. (1991). Une chaîne opératoire au Togo. Réflexions sur la méthode. In H. Balfet, (Ed.), Observer l’action technique: des chaînes opératoires, pour quoi faire? (pp. 65-86). Paris: Edition du Centre national de la recherche scientifique.

Ousterhout, R. G. (2017). Visualizing Community: Art, Material Culture, and Settlement in Byzantine Cappadocia. Washington D.C.: Dumbarton Oaks Research Library and Collection.

Öztürk, F. G. (2009). Kapadokya’da dünden bugüne kaya oymacılığı. Rock carving in Cappadocia from past to present. Istanbul: Arkeoloji ve Sanat Yayınları.

Öztürk, F. G. (2017). Rock-cut Architecture. In Ph. Niewöhner (Ed.), The archaeology of Byzantine Anatolia from the End of Late Antiquity until the Coming of the Turks (pp. 148-159). Oxford-New York: Oxford University Press.

Pearce, D. G. (2006). A Comment on Swart’s Rock Art Sequences and Use of the Harris Matrix in the Drakensberg: Notes and Comments. Southern African Humanities, 18(2.1), 173-188.

Pearce, D. G. (2010). The Harris Matrix Technique in the Construction of Relative Chronologies of Rock Paintings. South Africa. The South African Archaeological Bulletin, 65(192), 148-153.

Rouger, E. (1998). Du principe d’analyse stratigraphique à l’archéologie d’élévation: réflexion et méthode. Archéologie médiévale, 28, 161-172.

Russell, Th. (2000). The Application of the Harris Matrix to San Rock Art at Main Caves North, Kwazulu-Natal. The South African Archaeological Bulletin, 55(171), 60-70.

Russell, Th. (2012). No One Said It Would Be Easy. Ordering San Paintings Using the Harris Matrix: Dangerously Fallacious? A Reply to David Pearce. The South African Archaeological Bulletin, 67(196), 267-272.

Temel, A., Gündoğdu, M. N., Gourgaud, A., & Le Pennec, J.-L. (1998). Ignimbrites of Cappadocia Central Anatolia, Turkey: Petrology and Geochemistry. Journal of Volcanology and Geothermal Research, 85, 447-471.

Viereck-Goette, L., Lepetit, P., Gürel, A., Ganskow, G., Çopuroğlu, I., & Abratis, M. (2010). Revised Volcanostratigraphy of the Upper Miocene to Lower Pliocene Ürgüp Formation, Central Anatolian Volcanic Province, Turkey. The Geological Society of America Special Paper, 464, 85-112.

Wallace, S.-A. (1991). Byzantine Cappadocia: The Planning and Function of its Ecclesiastical Structures (Doctoral thesis). Australian National University, Canberra.

Haut de page

Notes

1 We refer to the methodological article by E. Rouger (1998) dealing with the application of Harris’s principles to archaeology of buildings.

2 The relevance of the method applied to rock art has been debated: Russell, 2000; Russell, 2012 contra Pearce, 2010. Method convincingly applied by the team working at the Lalibela site: Fauvelle et al., 2010.

3 I wish to thank B. Desachy for our discussions on the application of the software Le Stratifiant to rock cutting.

4 The dating of this ignimbrite layer is still being discussed: Le Pennec et al. 2005; Viereck-Goette et al. 2010.

5 For descriptive convenience, the apse is considered to face east.

6 This alteration was first noted by S.-A. Wallace (1991, II p. 532), who explains the irregularity of the covering by the presence of inclusion.

7 About these underground rooms, see Jolivet-Lévy & Lemaigre Demesnil, 2015, p. 33.

8 We have distinguished the picks according to the length of their handle because in Cappadocia, two variants of this tool have been observed. See Lamesa, 2011, p. 184; Lamesa, 2018, p. 79.

9 Bessac, 1986, pp. 14-24 for the long-handled pick or stonecutter pick.

10 On radial, slanted, and linear cutting: Bessac, 1986, pp. 19-20.

11 Additions to the table are indicated by the colour grey.

12 It should be noted that stone workers most often use a pneumatic hammer for hewing and cutting the bedrock.

13 J.-Cl. Bessac mentions the replacement of the squeegee with a very sharp chisel.

14 Additions to the table are indicated by the colour grey.

15 The team working at Lalibela applies this technique of reverse sequencing by studying the superimposition of impacts or ‘anomalies’ (in the words of the authors) on the site of the rock-hewn churches of Lalibela.

16 Wall measurements: 172 cm long, 214 cm high (measurement taken on the step) and 246 cm high (at the foot of the step).

17 On the question of unfinished works and sculptures, see the recent reflections of Vidya Dehejia & Peter Rockwell (2016).

18 On a first example of comparative work, see Lamesa & Gervers (forthcoming).

Haut de page

Table des illustrations

Titre Fig. 1: Two inclusions of decimetric and metric scale, the Church Göreme n°4b
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-1.jpg
Fichier image/jpeg, 133k
Titre Fig. 2: Plan of the Church Göreme n°4b
Crédits A. Lamesa and V. Dubois
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-2.jpg
Fichier image/jpeg, 42k
Titre Fig. 3: Details of the different elements composing the nave
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-3.jpg
Fichier image/jpeg, 92k
Titre Fig. 4: Bench’s irregularity to the northwest part of the nave with sledgehammer impacts
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-4.jpg
Fichier image/jpeg, 112k
Titre Fig. 5: View of the Church Göreme n°4b and its supposed entrance
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-5.jpg
Fichier image/jpeg, 103k
Titre Fig. 6: The apse and the southeast room of the Church Göreme n°4b’s nave
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-6.jpg
Fichier image/jpeg, 108k
Titre Fig. 7: View of the apse of the Church Göreme n°4b
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-7.jpg
Fichier image/jpeg, 159k
Titre Fig. 8: View of the unfinished room in the south part of the church
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-8.jpg
Fichier image/jpeg, 87k
Titre Fig. 9: The naves’ painting destroyed by a niche
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-9.jpg
Fichier image/jpeg, 239k
Titre Fig. 10: A long-handled pick and a short-handled pick held by stonemason K. Manap
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-10.jpg
Fichier image/jpeg, 286k
Titre Fig. 11: Long-handled picks’ impacts, west wall of the nave
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-11.jpg
Fichier image/jpeg, 197k
Titre Fig. 12: Polkas’ tool marks on the relief side on the east wall of the nave
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-12.jpg
Fichier image/jpeg, 82k
Titre Fig. 13: View of the east wall of the nave: junction between the short-handled picks’ tool marks and the polkas’ tool marks
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-13.jpg
Fichier image/jpeg, 179k
Titre Fig. 14: West wall of the nave of the Church Göreme n°4b, radial and slanted cutting by short-handled pick
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-14.jpg
Fichier image/jpeg, 203k
Titre Fig. 15: Inclusion of decimetric scale where the long-handled pick and sledgehammers’ tool mark appear
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-15.jpg
Fichier image/jpeg, 97k
Titre Fig. 16: Ceiling of the Church of St. Basil/Uçhisar n°3. Short-handled pick’s tool marks highlight the smoothing of the cross
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-16.jpg
Fichier image/jpeg, 295k
Titre Fig. 17: Using a rasp or an abrasive to smooth out the moulding of the T05 DB tomb located in the Davutlu Valley Bucağı at Mazıköy
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-17.jpg
Fichier image/jpeg, 154k
Titre Fig. 18: K. Manap using the polka as a scraper
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-18.jpg
Fichier image/jpeg, 93k
Titre Fig. 19: 1980s photograph of Mustafa Işık digging the depot near the Göreme Museum
Crédits Anonymous, given to A. Lamesa in 2008
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-19.jpg
Fichier image/jpeg, 61k
Titre Fig. 20: Delimitation of Negative Units
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-20.jpg
Fichier image/jpeg, 226k
Titre Fig. 21: Harris diagram of Negative Units taken from Le Stratifiant software, non-phased.
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-21.png
Fichier image/png, 51k
Titre Fig. 22: Harris diagram of the Negative Units taken from the software Le Stratifiant, phased
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-22.png
Fichier image/png, 48k
Titre Fig. 23: Inclusions close to the southeast room’s entrance
Crédits A. Lamesa
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/1308/img-23.jpg
Fichier image/jpeg, 182k
Haut de page

Pour citer cet article

Référence papier

Anaïs Lamesa, « Methods for Rock-hewn Worksite Analysis: The Church of Göreme n° 4b (Cappadocia): A Case Study »Anatolia Antiqua, XXVIII | 2020, 23-45.

Référence électronique

Anaïs Lamesa, « Methods for Rock-hewn Worksite Analysis: The Church of Göreme n° 4b (Cappadocia): A Case Study »Anatolia Antiqua [En ligne], XXVIII | 2020, mis en ligne le 01 décembre 2022, consulté le 12 mai 2025. URL : http://journals.openedition.org/anatoliaantiqua/1308 ; DOI : https://doi.org/10.4000/anatoliaantiqua.1308

Haut de page

Auteur

Anaïs Lamesa

post-doc fellow DIM-Matériaux anciens et patrimoniaux / UMR 8167 Orient et Méditerranée
External affiliated member at IFEA

Articles du même auteur

Haut de page

Droits d’auteur

CC-BY-SA-4.0

Le texte seul est utilisable sous licence CC BY-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search