Version classiqueVersion mobile

On salt, copper and gold

Catherine Marro
Thomas Stöllner

Early mining in the Caucasus and beyond

Salt‑exploitation techniques during the Early Bronze Age in the Caucasus

Technology and use-wear of the macro-lithic tools from Duzdağı and Kültepe I (Nakhchivan)

Caroline Hamon, Nicolas Gailhard et Gilles Fronteau


Le sel est une denrée à haute valeur ajoutée dans les pratiques rituelles et les réseaux d’échange comme le démontrent de nombreux exemples ethnographiques et préhistoriques. De récents travaux menés sur la mine de sel de Duzdağı (Nakhchivan) ont mis au jour une exploitation de sel gemme datée du Ve millénaire avant notre ère dans le sud du Caucase. Les vestiges archéologiques comprennent des centaines d’outils macrolithiques. L’analyse technologique et fonctionnelle des différents types de marteaux et outils de broyage issus des prospections nous conduit à proposer plusieurs hypothèses relatives à leur utilisation pour l’exploitation du sel, depuis son extraction jusqu’à sa transformation. Les outils d’extraction en particulier ont été utilisés à différentes étapes de la chaîne opératoire d’extraction, tandis que les meules et pilons attestent clairement de la transformation du sel sur le complexe minier lui-même.
L’étude des outils retrouvés dans les sondages réalisés entre 2011 et 2014 démontre un lien entre les types d’outils et les morphologies d’extraction propres à chaque période d’exploitation (âge du Bronze ancien et âge du Fer). En comparant les outils de Duzdağı avec les outils retrouvés sur le tell de Kültepe I, il a été possible de mettre en évidence un type de marteaux propres aux mineurs kuro-araxes. Il s’agit d’outils appointés réalisés sur une roche particulièrement résistante, un basalte alcalin, qui présentent un degré d’utilisation particulièrement poussé, un haut degré de recyclage et des durées d’utilisation importantes. Ces caractéristiques suggèrent que ces outils avaient un statut particulier, non seulement en matière de techniques d’exploitation, mais également pour les mineurs eux-mêmes. Ils interrogent le statut des mineurs de sel dans les sociétés kuro-araxes, et plus largement l’importance du phénomène minier dans les économies des populations de l’âge du Bronze caucasien.

Texte intégral

This study was carried out as part of the Mission archéologique du bassin de l’Araxe, hosted by the French Ministry of Foreign Affairs and directed by C. Marro, in collaboration with V. Bakhshaliyev from the Azerbaijan National Academy of Sciences at Nakhchivan. For the study of the collections from the site of Kültepe I, we would like to thank the director of the National Museum in Baku, the director of the Nakhchivan Museum and their respective teams for their warm welcome, as well as R. Berthon (Muséum national d’histoire naturelle, Archéozoologie et archéobotanique, UMR 7209).


1Salt is a highly valuable commodity in ritual practices and in exchange networks, as demonstrated in several ethnographic and prehistoric contexts (Harding 2013; Brigand, Weller 2015). In the South Caucasus, the prehistoric exploitation of rock salt is attested on the site of Duzdağı (Abibullayev 1982; Aliyev 1991), located 7 km from the modern city of Nakhchivan (Azerbaijan) (fig. 1). This salt dome, which appeared during the Middle/Late Miocene, towers above the left bank of the Aras River, between the north-western part of Iran and the eastern part of Turkey (Aali et al. 2012). At least three levels of very hard, pure salt, with a NaCl content of 86 to 98%, were exploited during Late Prehistory (Azizbekov 1961; Gonon et al., this volume).

Fig. 1 – Map showing the location of Duzdağı in Nakhchivan (MBA).

Fig. 1 – Map showing the location of Duzdağı in Nakhchivan (MBA).

2Recent archaeological investigations conducted by the French-Azerbaijani Araxes Research Program have revealed that salt exploitation was carried out at Duzdağı from the second half of the 5th millennium BC into the Middle Ages. Several survey campaigns (Marro, Bakhshaliyev, Sanz 2010) and various campaigns of test trenching carried out between 2007 and 2016 (Gonon et al., this volume) have provided a better overview of the different strategies of salt exploitation through time. These operations, together with a reassessment of the collections stored in the National Museum of History in Baku and in the Historical Museum of Nakhchivan, have provided hundreds of macro-lithic tools (hammers, grinding tools) that have been included in our study. Their technological and use-wear study make it possible to detail the long-term evolution of the exploitation processes and techniques employed at the Duzdağı salt mine.

3This study was focused on several issues, with a special emphasis on the dating of the different contexts and exploitation zones of the mine. Is there a link between the presence of hammers and specific exploitation zones? What kind of hammers were used in the salt-mining area, and for what purpose? Can we establish a correlation between the different types of hammer and the evolution of extraction systems brought to light at Duzdağı? Is it possible to identify the techniques and break down the chaînes opératoires used during different periods of exploitation? Can we identify an evolution in the exploitation techniques at Duzdağı? At a wider level, this study aims to assess the importance of salt mining in the economy and in the social structure of the populations who extracted, transformed and traded salt, especially during the Kura-Araxes period.

Methodology and classification of the macro-lithic tools

4The stone tools from three different contexts have been studied with a view to identifying their functions and characteristics:

  • the tools from the field surveys conducted between 2007 and 2011 (Hamon 2016);
  • the tools from the test trenches excavated between 2011 and 2016 by N. Gailhard and T. Gonon (Gonon et al., this volume);
  • the tools collected during the excavations of the Kura-Araxes levels of Kültepe 1 (Abibullayev 1982), a settlement located some 12 km to the north-east of the Duzdağı salt mine.

5Our analysis strategy for studying the Nakhchivani mining tools draws on the technological classifications from previous studies that have been carried out on the stone toolkits of other mining sites. It is based on an approach combining the study of raw materials, morphology, hafting systems and use-wear characteristics, with a particular focus on hammers.

6The general morphology, shaping techniques and hafting systems of the hammers have a major impact on their functioning. Their pointed shape and weight have important consequences on their efficiency for penetrating the salt or clay rock. Their hafting preparation and orientation (De Pascale 2004) correspond to different types of handles and to specific gestures. Their typology is based on the different kinds of notches made by flaking and pecking, and on the different forms of pecked grooves (simple or complex, symmetrical, peripheral or distal) (Garner 2013; Timberlake 2003; Zebrak 1995). These different criteria have been combined to propose a classification of the tools into different types.

7The approach was extended to include use-wear analysis, which is essential for determining the type of targeted material, as well as the cycles and intensity of use of these tools (Figuls et al. 2007; Timberlake, Craddock 2013; Weller, Figuls, Grandia 2007). Use-wear analysis was conducted at low magnification with a Nikon SMZ800 binocular microscope (up to 63x), following a methodology developed for macro-lithic tools in general (Hamon 2008; Adams et al. 2009).

Stone tools from the field survey

The context

8Between 2008 and 2011, four campaigns of an extensive walk-over survey were carried out on the southern and western slopes of the salt mountain; they resulted in the creation of 6 test zones. In these zones, a programme of intensive recording and georeferenced plotting of all pottery sherds, obsidian artefacts and macro-lithic tools was conducted directly in the field by using a geographic information system (ArcGIS) combined with a DGPS system (Marro, Bakhshaliyev, Sanz 2010). In the test zones selected, large quantities of Kura-Araxes ceramics attest to a significant period of exploitation that occurred during the Early Bronze Age. A small number of pottery sherds dating to the Iron Age and the medieval period have also been collected and testify to a later exploitation of the salt deposit (Marro, this volume). Up to 718 hammers have been recorded and studied during these combined field and lab operations. Their detailed study has provided the main descriptive and interpretative framework for identifying their function (Hamon 2016). The main results of this work are summarised below.

The tools

9Most of the raw materials used for making the hammer tools were collected from the old terraces of the Araxes river, which are still visible on top of the summit of the Duzdağı salt mountain. A large diversity of rocks was selected to craft the hammers. The rock types identified include sedimentary rocks, especially sandstones and limestones, as well as volcanic rocks such as alkali and vesicular basalts. A few granites and metamorphic rocks were also selected occasionally. From a mechanical and functional perspective, hammers are made of both soft and hard stones, which clearly do not offer the same range of functional properties.

10Eight categories of hammers have been defined (tab. 1; fig. 27):

  • Type A (7.9% of hammers) appears as simple cobbles, the ends of which display splintering and bruising marks resulting from their use. The opportunistic selection of the blanks, as well as the rather short-duration use of these tools, suggest that they may be regarded as occasional hammers that were probably used for secondary or low-skilled tasks.
  • Type B (6.7%) features simple central or distal lateral notches that represent the simplest type of hafting system. The removal of one or two flakes from the sides of the pieces is the most frequent technique used to create these notches, although pecking is also used.
  • Type C (13.5%) features simple hafting grooves; these hammers are made from very diverse raw materials: the size of the blanks also varies. Hammers of this type have one or two central or distal grooves to facilitate hafting. These grooves are generally symmetrical and located on the faces or sides.
  • Type D (7.2%) combines at least two grooves: one on the face or sides of the hammer, and another at the extremity; these grooves would have facilitated a complex hafting system.
  • Type E (46.2%) were manufactured from alkali basalt, a particularly hard and resilient material. These pieces exhibit distinctive pointed distal ends, intensively shaped by pecking. They bear evidence of intensive use cycles and encompass a wide range of weights.
  • Type F (9.6%) have particularly significant weights, ranging from 5 to 80 kg. Several subtypes can be distinguished: quadrangular-shaped hammers, hammers with distal lateral grooves for facilitating hafting, etc.

Tab. 1 – Macro-lithic tools from Duzdağı, survey campaign 2008‑2011.

Hammers   718
  A non-manufactured tools 57
  B notched tools 48
  C simple grooved tools 97
  D complex grooved tools 52
  E basalt tools 332
  F massive tools 69
  G other 61
  H perforated 2
Grinding slabs/Handstones   10
Hammerstones   50
Intermediate tools   6
Pestles   30
Others   6
Total   820

Fig. 2 – Duzdağı, survey, hammer types A and B (C. Hamon).

Fig. 2 – Duzdağı, survey, hammer types A and B (C. Hamon).

Fig. 3 – Duzdağı, survey, hammer types C and D (C. Hamon).

Fig. 3 – Duzdağı, survey, hammer types C and D (C. Hamon).

Fig. 4 – Duzdağı, survey, hammer type E (C. Hamon).

Fig. 4 – Duzdağı, survey, hammer type E (C. Hamon).

Fig. 5 – Duzdağı, survey, stone tools type F (C. Hamon).

Fig. 5 – Duzdağı, survey, stone tools type F (C. Hamon).

Fig. 6 – Duzdağı, survey, diverse hammer types (C. Hamon).

Fig. 6 – Duzdağı, survey, diverse hammer types (C. Hamon).

Fig. 7 – Duzdağı, survey, grinding tools; a‑b: quern; c: grinder; d: pestles (C. Hamon).

Fig. 7 – Duzdağı, survey, grinding tools; a‑b: quern; c: grinder; d: pestles (C. Hamon).

11Together with these hammer types, several grinding and pounding tools, essentially querns and pestles, were also recorded during the survey. Their characteristics clearly match tools of the same category found on prehistoric settlements. The querns are made of vesicular basalt and the pestles are made of truncated-cone shaped cobbles.


12On the basis of these observations and the use-wear analysis of a 287 tool sample (35% of the series) three functional groups of hammers can be proposed.

Intermediate size hammers (Types A to D)

13Intermediate size hammers (Types A to D) may have had multiple functions. This multi-functionality is suggested by the low investment in the choice of raw material and in the shaping of their distinctive morphology. The use of soft or semi-soft rocks, with low cohesion, suggests that these tools are not suitable for direct extractive operations on dry rock salt but may have been occasionally used on altered or wet salt. Alternatively, they may have been used to remove the clay layers that overlie the salt deposits, or, perhaps, to shape the salt blocks after their extraction, or both. To sum up, most of these hammers may have been used for non-specialized tasks and multiple operations connected with salt extraction.

Massive hammers (Type F)

14Because of their weight, and the specific constraints this implies, massive hammers (Type F) must have had a specific role in the extractive sequence. The most massive hammers may have been held in rope-slings in order to hurl them against the rock face with a swinging motion. It is also possible that they were swung from tripods, as was the case at other mines in the Near East. Two particular operations could have required their specific use: they could have been used either for breaking and removing the limestone slabs located just above the salt layers, or for opening tunnels and shafts.

Type E hammers

15Type E hammers clearly had a specific function and status within the chaîne opératoire of salt extraction. It is the only category of hammer that occurs in different sizes and weights. All the tools belonging to this category have been made from a particularly shock-resistant rock, alkali basalt, possibly originating from a non-local source located up to 10 km to the north of the Duzdağı salt mine. The petrographic analyses indicate that they were made from basalt or a related magmatic rock, with a plagioclase-phyric texture where the mesostase is largely recrystallized (fig. 8). Altered plagioclases phenocrystals, sometimes zoned and with a length of more than 2 mm, are embedded in a substantial mesostasis, mainly composed by chlorite and calcite. In the case of the Duzdağı tools under study, the recrystallization is considerable and leaves some doubt about the real primary texture of the magmatic rock: true volcanic texture or subvolcanic texture? But the microfacies is very similar to the one that has been identified as a southern Armenian basanite in Galoyan et al. (2009). In addition, it must be noted that chlorite crystallisations are also observed in the plagioclase phenocrystals alteration features, resulting in a solid, cohesive, non-porous stone. The properties of this particularly resistant rock would have rendered it efficient for the extraction of rock salt and would have favoured a long duration use. Furthermore, their very well-shaped sharp point is particularly appropriate and effective for the fine outlining and extraction of salt blocks.

Fig. 8 – Duzdağı survey, microphotographs of the thin section taken from a Type E hammer (no. 445) from Duzdağı; a‑b: phyric texture with large plagioclase crystals in a homogeneous recrystallized mesostase, scale bar 1 mm; c‑d: chlorite-rich mesostase with some epidote and calcite crystals, scale bar 200 µm; a and c under parallel nichols; b and d under crossed nichols (G. Fronteau).

Fig. 8 – Duzdağı survey, microphotographs of the thin section taken from a Type E hammer (no. 445) from Duzdağı; a‑b: phyric texture with large plagioclase crystals in a homogeneous recrystallized mesostase, scale bar 1 mm; c‑d: chlorite-rich mesostase with some epidote and calcite crystals, scale bar 200 µm; a and c under parallel nichols; b and d under crossed nichols (G. Fronteau).

16Finally, querns and pestles complete the overview of the technical stages of salt exploitation. Judging from the use-wear analysis, these tools were brought over from prehistoric settlement sites to be reused at the Duzdağı salt mine for crushing and grinding salt. The low number of these tools suggests that the on-site grinding of salt was an occasional activity, with small quantities of salt processed, probably by small, local, or even family, groups.

17This overview of the characteristics of the tools recovered during the field survey clearly stresses the importance of stone tools in the salt-extraction techniques used by the Kura-Araxes miners of Duzdağı. It highlights the contrast between less specialized tools, which were probably used for a broad spectrum of activities, and a more specialized category of hammers, namely the pointed tools, which were particularly effective for salt extraction.

The stone tools from test trenches M1 and M11

18Several test trenches have been excavated on the south-western slopes of the Duzdağı salt mountain in order to shed light on the layout of the extraction sites and working areas (Gonon et al., this volume). These test trenches were situated in areas where concentrations of Late Chalcolithic, Kura-Araxes and Iron Age pottery were identified during the survey campaigns, which suggests intensive exploitation activity. From the ceramic potsherds found in these areas, it appears that the lower part of W2 (Gonon et al., this volume) was exploited during the Kura-Araxes period and the Iron Age (Marro, this volume). No clear mining infrastructure could be related to the Late Chalcolithic finds, which suggests that salt exploitation at this time was occasional and opportunistic. Exploitation by Kura-Araxes groups, on the other hand, is characterised by the existence of mining cells (M1), i.e. small work areas, probably excavated by a single miner, which are possibly related to a small-scale salt procurement system. Trench M11 testifies to the existence of platforms close to the M1 mining operation where a Kura-Araxes jar was found together with stone tools and a hearth. The ubiquity of the association of Kura-Araxes potsherds and stone tools in this zone and all over the surveyed areas, however, testifies to the importance of salt in the Kura-Araxes economy.

19In trench M1, a second phase of extraction has been identified in the Late Bronze Age or Early Iron Age. The work area opened by Iron Age miners was again an opencast but a new technique of extraction was used: salt was exploited by cutting large slabs out of the salt bed. The cutting of sophisticated products, such as slabs, from the salt deposits suggests that salt had acquired an added, possibly prestigious, value during the Late Bronze/Early Iron Age (Gonon et al., this volume).

Early Bronze Age exploitation: trenches M1 and M11

The context

20During the field survey, a concentration of Kura-Araxes sherds and macro-lithic tools (Window 2) was recorded over a wide area that includes large platforms. In this sector, two test trenches were opened: one at the base of a salt outcrop (M1) and another on a fairly level area a little further away to the south-east (M11).

21Test trench M1 has yielded two extraction zones with very different morphologies. On the face of the salt outcrop, several pits and mining cells have been discovered (fig. 9). This type of morphology clearly corresponds to a small-scale exploitation area, each cell corresponding to a single area of extraction, which could accommodate a single person in a crouching or kneeling position.

Fig. 9 – Bronze Age mining cells in trench M1 (MBA, N. Gailhard).

Fig. 9 – Bronze Age mining cells in trench M1 (MBA, N. Gailhard).

22Test trench M11 is located close to the south-east of test trench M1, on a platform where one of the largest concentrations of Kura-Araxes ceramics and macro-lithic tools was identified. In trench M11, we immediately reached the Kura-Araxes occupation layers. These layers were not very thick because of significant erosion but we could make out three different work layers. In the earliest, an almost complete jar, typical of the Kura-Araxes period, was found. This jar appears as a good example of the containers used on the mine, since the majority of diagnostic sherds come from vessels of the same type. These are narrow-necked jars (between 13 and 16 cm in diameter), coarse-grained, usually beige or chamois in colour, with “Nakhchivan” type handles. Similar ethnographic examples show that the three handles attested on these jars may have been used to fix a rope in order to carry large volumes of water (Marro, this volume). In trench M11, it is remarkable to note that the entire assemblage, i.e. a total of 1,037 sherds, consists exclusively of Kura-Araxes ceramics.

The tools

23The macro-lithic tool assemblages from the EBA exploitation of trenches M1 and M11 are very similar (tab. 2). They include hammers and chopping tools, probably related to salt extraction, as well as other categories such as hammer stones (only found in M1) used for the shaping of tools and, in both areas, pestles related to the grinding of the rock salt on site (fig. 10b).

Tab. 2 – Macro-lithic tools from the test trenches (2011‑2015).

Tab. 2 – Macro-lithic tools from the test trenches (2011‑2015).

Fig. 10 – Duzdağı excavations, Kura-Araxes trench M1; a: intermediate tool; b: pestle; c‑e: hammers; f: polisher (C. Hamon).

Fig. 10 – Duzdağı excavations, Kura-Araxes trench M1; a: intermediate tool; b: pestle; c‑e: hammers; f: polisher (C. Hamon).

24Four different categories of stone artefacts have been discovered within this trench:

  • tools for salt extraction; six hammers belonging to three standard types, among which Type D hammers (characterized by complex hafting grooves), pointed Type E hammers, and massive Type F hammers (tab. 3); the type E hammers are clearly at an advanced stage of use: they are very small in size, their ends have been completely smashed and blunted, while the hafting grooves have been damaged by secondary flaking; these hammer types all correspond to fairly specialized types of tools that have been used in specific stages of the chaîne opératoire during salt extraction;
  • two chopping tools made out of natural sandstone cobbles; these display a well-defined sharp edge created by the removal of large flakes, with intense percussion marks, and on the opposite end a concentration of fine percussion impacts; they were clearly used as intermediate tools;
  • tools for making tools; the shaping of tools on the site is evidenced by the discovery of a couple of hammer stones, several natural cobbles, shaping flakes and hammer fragments; this evidence mainly relates to the manufacture of pointed hammers; in addition, several chopping tools with intense percussion marks have been recovered in this exploitation area;
  • tools for processing salt; small truncated-cone shaped pestles, about 10 cm in length, which were made from basalt or sandstone cobbles; some examples bear dense, regular pecking on their surface; they display at least one large plano-convex distal end, covered with fine percussion marks and complete smoothing of the surface; a second active surface sometimes occurs at the opposite end; limited percussion zones are found on the sides and faces of the tool; they attest to salt-grinding activity around the exploitation zone.

Tab. 3 – Hammer types from the test trenches (2011‑2015).

Tab. 3 – Hammer types from the test trenches (2011‑2015).

25An unusual tool rough-out found in the same levels is of special interest. This tool is made of metamorphic limestone, displaying a back that has been entirely evened flat by pecking. One of its narrower ends has been shaped by flaking, so as to prepare a “handling” area. Its second part, which is circular in section, has been delimited by a series of flakes around a central, slightly concave, circular pecked area. This rough-out does not resemble a pounding or hammering tool, and its use remains unknown.

26To sum up, the tools from the EBA test trenches of Duzdağı have yielded quite a varied lithic toolkit, which includes extraction hammers and intermediate tools, workshop tools and finally grinding tools.


27Almost all the categories of tools found during the field survey are represented in the Early Bronze Age assemblage from zone M1. The three functional groups of hammers are represented, from the massive digging tools to the more multifunctional tools, although the majority of the assemblage is composed of Type E pointed tools. They all show a certain crafting investment. They reflect the entire process of salt transformation, from the shaping of hammers in the work areas located near the extraction zones to their actual use, specialized or not, for extracting and grinding salt.

Late Bronze/Early Iron Age exploitation

The context

28To the north of the Kura-Araxes exploitation zone, a second extraction area has been brought to light: it is characterised by the horizontal extraction of quadrangular slabs of salt (fig. 11). This area opens directly onto the salt bed, thus avoiding the need for underground digging. Pottery dating to the Late Bronze/Early Iron Age was collected, while 14C dating was conducted on wood and other organic remains. One reliable 14C date can be connected directly to the slab-extraction zone: it was obtained from a hearth (US 45) located a short distance to the west; it was dated to between 1200 and 1000 cal. BC (Gonon et al., this volume). These finds suggest that the extraction techniques and the chaîne opératoire differed greatly between the Early Bronze age and the Late Bronze/Early Iron Age.

Fig. 11 – Iron Age slab‑extraction area from trench M1 (MBA, N. Gailhard).

Fig. 11 – Iron Age slab‑extraction area from trench M1 (MBA, N. Gailhard).

The tools

29A dozen tools were retrieved from the Late Bronze/Early Iron Age exploitation zone. Only a few hammers were recovered from that area (fig. 12). They are generally made from natural cobbles, without any hafting preparation. Only one massive hammer (Type F, fig. 12f) displays two pecked grooves on the side that were prepared for hafting, and limited percussion marks on the distal end. Another hammer made from a cobble shows one narrow hammered end and, at the other end, a wide percussive surface with flaking.

Fig. 12 – Duzdağı excavations, Iron Age tools from M1; a: possible pick; b: intermediate tool; c‑d: pestles; e: grinding tool; f: massive hammer (C. Hamon).

Fig. 12 – Duzdağı excavations, Iron Age tools from M1; a: possible pick; b: intermediate tool; c‑d: pestles; e: grinding tool; f: massive hammer (C. Hamon).

30Four pestles have been found (fig. 12c‑d), including two probable rough-outs: their presence in the extraction area may indicate that such tools were produced on site in work areas. One truncated-cone shaped pestle was manufactured from the fragment of a pointed hammer. Another cylindrical pestle displays broad, convex percussive ends. Several fragments of cobbles also show percussion marks on their ends or sharp edges.

31A couple of block fragments have clearly been used, or reused, for percussion actions. Some of them appear to be intermediate tools, as they display a totally flaked narrow distal end and a large hammered area at the proximal end.

32To sum up, very few hammers have been found in this area of exploitation, and none of them correspond to the pointed Type E. Most of the Iron Age toolkit is composed of intermediate tools (found in very small numbers during the surveys) and pestles used for the grinding of salt.


33The toolkit from the Late Bronze Age/Early Iron Age area clearly corresponds to an opportunistic tool production, with a very low investment in the selection and shaping of the natural blanks, especially in the case of hammers. We may note a complete absence of complex hammers (Type D) and pointed hammers (Type E), which make the bulk of the survey assemblage and the areas with Early Bronze Age ceramic concentrations. These observations indicate a very low degree of specialization of the hammers and percussion tools from the LBA/EIA extraction zone. These non-specialized hammers and percussion tools may have been used to outline and detach the salt slabs from the outcrop. They may have been used to complete a broader toolkit that also included wooden tools, such as wedges for example (Gonon et al., this volume).

34More surprising is the presence of several pestles in this area. Unless these tools were involved in the shaping of the slabs, which is an unlikely hypothesis given their characteristics, they may indicate the existence of a work place for salt transformation associated with this extraction zone.

35It is clear that the stone toolkit from the Late Bronze Age/Iron Age area is very distinct from the assemblage found in Early Bronze Age exploitation areas: it differs by its opportunistic character, the presence of intermediate tools and the absence of pointed Type E hammers.

The stone tools from Kültepe I

The context

36The site of Kültepe I is located 12 km to the east of Duzdağı, on the left bank of the Naxcivan Çay, a northern tributary of the Araxes River. It is one of the main settlements of the region and was occupied as early as the Neolithic. In his publication of the Kura-Araxes excavations, Abibullayev (1982) presented a series of hammers whose morphology clearly recalls that of the Duzdağı tools (fig. 13). A reassessment of the entire lithic assemblage from these excavations has been carried out by C. Hamon at the National Museum of History in Baku, within the framework of a Shelby‑White foundation project conducted by R. Berthon (Muséum national d’histoire naturelle, Archéozoologie et archéobotanique, UMR 7209).

Fig. 13 – Kültepe I, stone tools from the Kura-Araxes layers (after Abibullayev 1982); 6‑8 are possible hammers; 13‑14 are possible pestles (C. Hamon).

Fig. 13 – Kültepe I, stone tools from the Kura-Araxes layers (after Abibullayev 1982); 6‑8 are possible hammers; 13‑14 are possible pestles (C. Hamon).

The tools

37As well as domestic tools such as querns, mortars and percussion tools, about fifty hammers have been recovered; they all show direct comparisons with the tools found in Duzdağı (fig. 14). Two thirds of this series are composed of pointed Type E hammers. They are made of the same very hard basaltic rock as those from Duzdağı and display a very fine shaping resulting from over-all surface pecking. Their small dimensions suggest that the individual hammers have been used intensively over a long period of time; some examples even show evidence of re-hafting (double grooves). Together with this dominant hammer type, eleven simple non-shaped hammers (Type C) display peripheral or distal grooves. Two massive Type F hammers (one more than 30 cm long) were also collected.

Fig. 14 – Kültepe I, hammer types belonging mainly to type E (collections of the National Museum of History in Baku, photos: C. Hamon).

Fig. 14 – Kültepe I, hammer types belonging mainly to type E (collections of the National Museum of History in Baku, photos: C. Hamon).


38The re-examination of the macro-lithic tools from the Kura-Araxes layers of Kültepe has clearly revealed the presence of hammers intended for mining activities. A large majority of them find clear parallels with the assemblage from Duzdağı. This direct comparison has two implications. First, the abundant pointed hammers dating to the Early Bronze Age levels are essentially related to the Kura-Araxes exploitation of the salt mine. Secondly, we need to examine the possibility that the inhabitants of Kültepe I played a central role in the exploitation of the Duzdağı salt levels during the Early Bronze Age. One must note that other sites, such as Kültepe II, may also have exploited the salt from Duzdağı.

39To conclude, it is worth noting that no stone tools have been found in trench M9 where a Late Iron Age/Early Antiquity exploitation was identified, suggesting the complete abandonment of this kind of toolkit at this time.


Dating of the tools

40Since the Duzdağı mining zone has been exploited over a very long period of time, the precise dating of the different extraction zones is sometimes difficult to establish, at least on the basis of the information gathered from the field survey. The dating of each salt-extraction and exploitation zone is principally based on ceramic concentrations and on a number of dates provided by organic material from the different test trenches (Gonon et al., this volume). The survey has shown that some areas have been more thoroughly invested by Kura-Araxes miners during the Early Bronze Age. The data obtained from the test trenches and the excavations gave ample evidence that the salt-extraction strategies developed in these two periods widely differed (Gonon et al., this volume). Different and specific toolkits have been found in association with each form of extraction. In fact, while some non-specific types of hammers have been found in both Kura-Araxes and Iron Age contexts, only the Kura-Araxes miners seem to have produced and used the complex and pointed Type E hammers. This interpretation is confirmed by the discovery of quite a large number of such hammers in the Kura-Araxes occupation layers of Kültepe I in the Soviet excavations.

Of tools and techniques: a specialized activity?

41The study of the toolkit yielded by the field survey has already highlighted the relationship between the different kinds of tools and their role in the different stages of exploitation. The functional analysis has led us to propose three different stages of exploitation that would require specific tools: massive hammers for digging into the outcrops, pointed hammers for outlining and extracting salt blocks, and finally pounding tools for crushing the salt into powder. However, by comparing the hammers from the field survey and from the excavation contexts we have gained a better understanding of the function of these tools in the exploitation strategies of the different periods.

42Simple-stone hammers have been found within the precinct of the slab-extraction zone. In the Late Bronze/Early Iron Age extraction strategy, stone tools were not central to the technical process, nor do they seem to have contributed to specialized operations within the chaîne opératoire. Instead, they were probably used on a short-term, opportunistic basis as supplementary tools, with an exception perhaps for the intermediate tools. These tools were probably used in indirect percussion with other percussive tools (made of wood for example) for the detachment of salt slabs. This technique has also been identified on other Iron Age mining sites in north-western Iran (Aali et al. 2012).

43By contrast, the pointed Type E hammers seem to have played a major part in the extraction techniques employed by Kura-Araxes miners. Their quantity, long duration of use and particularly resistant raw material clearly suggest that they were specialized tools within the extraction process, probably used for the detachment and shaping of the blocks. In fact, they are associated with a specific extraction technique operating from mining cells. This suggests that the use of these tools is closely correlated to the morphology of the extraction locus: in the context of a small-scale extraction zone, the cutting out of blocks located next to others requires the fine and precise outlining of the blocks so as to save as much raw material and energy as possible during the extraction process. Presuming that the miners were producing their own tools, we can envisage at least two standardized stages in the extraction process: the shaping of the hammers and their use for the extraction of the salt blocks. The presence of very few grinding tools could be rather related to opportunist and small episodes of salt processing directly in the mine, for domestic or direct consumption (Hamon 2016).

44This also has implications for the nature, organization and intensity of the activities that took place in the mine. Although we are clearly looking at small-scale Early Bronze Age exploitations, the widespread presence of Kura-Araxes ceramics over the salt dome suggests repetitive episodes of extraction activity. In a first hypothesis, the exploitation episodes may have been quite close together in time; they would have mobilized a large part of the community, sharing a common know-how. In a second hypothesis, these exploitation episodes may have been much more spread out in time; this would imply know-how and knowledge transfer from one generation to the next. In the second hypothesis, it is more likely that the quarrying work was controlled and carried out by a specialized group within the community.

Symbolic and identity-defining tools?

45The use of a specialized and emblematic category of hammer suggests a desire on the part of the miners themselves, or of the community to which they belong, to express and materialize a mining component in the Kura-Araxian identity (Sagona 2014). As it has no functional necessity, the particular investment put in the shaping and finishing of these tools could be interpreted as a way to distinguish their know-how and to identify its special status at least within the community of miners. This mining identity may relate either to their mastering of a specific skill in the mining technique, or more specifically to the exploitation of salt, from extraction to trade.

46While grooved hammers are widespread within mining contexts, regardless of their date (non-metal using populations) or geographical location, they generally display a low degree of shaping and standardization, particularly focused on the basic preparation of hafting grooves (Craddock, Lang 2003; Delgado Raack, Escanilla Artigas, Risch 2014; Garner 2013; De Pascale 2004; Pickin 1990). Similar hammers and picks, characterized by their pointed ends generally shaped by flaking and pecking, are recorded from a small number of copper-mining sites: examples include granodiorite and cornean picks from the copper mine of Solana del Bepo, which is dated to the end of the 3rd millennium BC (Rafel et al. 2016), or the hammers from the Grotta della Monaca (Calabria, Italy; Breglia, Caricola, Laroca 2016). In the South Caucasus, similar pointed tools are not commonly found on other Bronze Age mining sites. None has been found in association with the well-known copper and gold extraction sites at Sakdrisi (Georgia) and in the Racha region (Stöllner et al. 2014; Tamazashvili 2014). This kind of tool is not as frequently found in the miners’ toolkit as might be expected, given that it is not used solely in salt exploitation.

47The second hypothesis to explore is that these were specialist tools for the extraction of salt. Yet again, very few parallels can be found in the toolkits from other Bronze Age salt mines to answer this question. For example, the toolkit presumed to be associated with the Cardona Neolithic salt mine (Catalunya, Spain) clearly displays very different characteristics: together with pestles, the supposed extraction tools are mainly axes (Figuls et al. 2007; Weller, Figuls, Grandia 2007). In the late Bronze Age exploitation of the Hallstatt mine (Austria), “the extraction technique was by bronze pics to create deep parallel grooves in the rock salt surface, and then to hammer the intervening rock out in lumps or chips” (Harding 2013, p. 64, after Kern et al. 2009). In fact, the possibilities of comparisons with other Bronze Age salt mines are too rare to investigate the specific relationship of these tools with salt mines.

48Actually, the Duzdağı hammers do not find any direct parallels in the Early Bronze Age copper mines of the Caucasus or elsewhere, and only a handful of sites have yielded implements that are a bit similar with pointed ends shaped by particularly fine pecking. They do not seem to correspond to a specific adaptation for salt-extraction techniques. However, they can be seen as prefiguring, in a certain way, the pointed metal picks found in later salt exploitation contexts, especially during the Iron Age (see for example Stöllner, Schwab 2009).

49The pointed hammers from Duzdağı, characterized by fine shaping and their distinctive pointed ends, do not find direct parallels in any related contemporary or functional contexts. They could be considered as a true technical innovation within the salt-extraction process but may also represent a desire on the part of the Duzdağı salt miners to express the special status of their activity within the community.


50The exploitation of salt in Duzdağı can be interpreted as a specialized and possibly valorizing activity among the Kura-Araxes people during the Early Bronze Age in the southern Caucasus. The high number of extractive tools on the surface of this salt dome suggests intensive exploitation during the Early Bronze Age. The presence of these tools not only in the mine but also on two of the most important Kura-Araxian settlements of the region, Kültepe I and Kültepe II, suggests a close relationship between their inhabitants and the quarrying activity at the salt mine. In this sense, we may surmise the existence of a mining complex, exploited over a long period of time by the Kura-Araxes populations who inhabited the region. If the mining phenomenon contributed to the economic development of the Kura-Araxes communities, it may also have played a structuring social role in the building of their identity, as expressed by the particular care and investment dedicated to their emblematic stone tools.


Aali et al. 2012: A. Aali, T. Stöllner, A. Abar, F. Rühli, “The salt men of Iran. The salt mine of Douzlakh, Chehrabad”, Archäologisches Korrespondenzblatt 42/1, 2012, pp. 61‑81.

Abibullayev 1982: O.A. Abibullayev, Eneolit i Bronza na Territorii Nakhichvanskoi ASSR, Baku, Elm, 1982.

Adams et al. 2009: J. Adams, S. Delgado, L. Dubreuil, C. Hamon, H. Plisson, R. Risch, “Functional analysis of macro-lithic artefacts”, in F. Sternke, L. Eigeland, L. Costa (dir.), Non-flint raw material use in Prehistory. Old prejudices and new direction. 15th UISPP Congress (Lisbon, September 2006), BAR International Series 1939, Oxford, Archaeopress, 2009.

Aliyev 1991: V.G. Aliyev, Kul’tura Epokhi Srednei Bronzy Azerbaidzhana, Baku, Elm, 1991.

Azizbekov 1961: C.A. Azizbekov, Geologia Nakhichvanskoi ASSR. Gosydartstvenoe naytchno tekhnitcheckoe izdatelstvo literatyre po geologiniox ranenedr, Moscow, 1961.

Breglia, Caricola, Laroca 2016: F. Breglia, I. Caricola, F. Laroca, “Macro-lithic tools for mining and primary processing of metal ores from the site of Grotta della Monaca (Calabria, Italy)”, Journal of Lithic Studies 3/3, 2016, pp. 2‑19.

Brigand, Weller 2015: R. Brigand, O. Weller, Archaeology of salt. Approaching an invisible past, Leiden, Sidestone Press, 2015.

Craddock, Lang 2003: P. Craddock, J. Lang, Mining and metal production through the ages, London, British Museum Press, 2003.

Delgado Raack, Escanilla Artigas, Risch 2014: S. Delgado Raack, N. Escanilla Artigas, R. Risch, “Mazas ocultas. Rastros de minería prehistórica en el cerro minado de Huercal-Overa (Almería)”, Cuadernos de Prehistoria y Arqueología de la Universidad de Granada 24, 2014, pp. 13‑44.

De Pascale 2004: A. De Pascale, “Hammers from early copper mines: sintesi dei ritrovamenti nell’Europa e nel Mediterraneo orientale e prime considerazioni sui mazzuoli di Monte Loreto (IV millennio BC-Liguria)”, Rivista di StudiLiguri 69, 2004, pp. 5‑42.

Figuls et al. 2007: A. Figuls, J. Bonache, J. Aranda, J. Sune, M. Vendrell, J. Gonzalez, A. Funetes, J. Mata-Perello, J. Sanz, “Neolithic exploitation of halite at the ‘Vall Salina’ of Cardona (Catalonia, Spain)”, in A. Figuls, O. Weller (dir.), Trobada International d’Arqueologia envers l’explotació de la sal a la Prehistória i Protohistória. Acts of the 1st International Archaeology meeting about Prehistoric and Protohistoric salt exploitation (Cardona, 6-8 dec. 2003), Cardona, IREC, 2007, pp. 119‑218.

Galoyan et al. 2009: G. Galoyan, Y. Rolland, M. Sosson, M. Corsini, S. Billo, C. Verati, R. Melkonyan, “Geology, geochemistry and 40Ar/39Ar dating of Sevan ophiolites (Lesser Caucasus, Armenia). Evidence for Jurassic back-arc opening and hot spot event between the South Armenian Block and Eurasia”, Journal of Asian Earth Sciences 34, 2009, pp. 135‑153.

Garner 2013: J. Garner, Das Zinn der Bronzezeit in Mittelasien II. Die montanarchäologischen Forschungen an den Zinnlagerstätten, Mainz, Philipp von Zabern, 2013.

Hamon 2008: C. Hamon, “Functional analysis of stone grinding and polishing tools from the earliest Neolithic of north-western Europe”, Journal of Archaeological Science 35, 2008, pp. 1502‑1520.

Hamon 2016: C. Hamon, “Salt mining tools and techniques from Duzdağı (Nakhchivan, Azerbaijan) in the 5th to 3rd millennium BC”, Journal of field archaeology 41/4, 2016, pp. 510‑528.

Harding 2013: A. Harding, Salt in Prehistoric Europe, Leiden, Sidestone Press, 2013.

Kern et al. 2009: A. Kern, K. Kowarik, A.W. Rausch, H. Reschreiter, Kingdom of Salt. 7000 years Hallstatt, VPA 2, Vienna, Natural History Museum, 2009.

Marro, Bakhshaliyev, Sanz 2010: C. Marro, V. Bakhshaliyev, S. Sanz, “Archaeological investigations on the salt mine of Duzdağı (Nakhchivan, Azerbaijan)”, TUBA-AR 13, 2010, pp. 229‑244.

Pickin 1990: J. Pickin, “Stone tools and early metal mining in England and Wales”, in P. Crewe, S. Crew (dir.), Early mining in the British Isles. The proceedings of the early mining workshop at Plas Tan y Bwlch, Snowdonia National Park Study Centre (17‑19 Nov. 1989), published in Plas Tan y Bwlch Occasional Paper 1, 1990, pp. 39‑42.

Rafel et al. 2016: N. Rafel, I. Montero Ruiz, I. Soriano, S. Delgado-Raack, “L’activité minière préhistorique dans le nord-est de la péninsule ibérique. Étude sur la Coveta de l’Heura et l’exploitation du cuivre à la Solana del Bepo (Tarragone, Espagne)”, Bulletin de la Société préhistorique française 113/1, 2016, pp. 95‑130.

Sagona 2014: A. Sagona, “Rethinking the Kura-Araxes Genesis”, Paleorient 40/2, 2014, pp. 131‑154.

Stöllner et al. 2014: T. Stöllner, B. Craddock, I. Gambaschidze, G. Gogotchuri, A. Hauptmann, A. Hornschuch, F. Klein, I. Löffler, G. Mindiashvili, B. Murwanidze, S. Senczek, M. Schaich, G. Steffens, K. Tamasashvili, S. Timberlake, “Gold in the Caucasus. New research on gold extraction in the Kura-Araxes Culture of the 4th millennium BC and early 3rd millennium BC”, in H. Meller, E. Pernicka, R. Risch (dir.), Metalle der Macht. Frühes Gold und Silber, published in Tagungen des Landesmuseums für Vorgeschichte Halle 11, 2014, pp. 71‑110.

Stöllner, Schwab 2009: T. Stöllner, R. Schwab, “Hart oder weich? Worauf es ankommt. Pickel aus dem prähistorischen Bergbau in den Ostalpen. Festschrift für F.E. Barth”, Mitteilungen Anthropologische Gesellschaft Wien 139, 2009, pp. 149‑166.

Tamazashvili 2014: K. Tamazashvili, Prehistoric mining implements from Georgia, Tbilisi, Mtsignobari, 2014.

Timberlake 2003: S. Timberlake, “Early mining research in Britain: the developments of the last ten years”, in P. Craddock, J. Lang, Mining and metal production through the ages, London, British Museum Press, 2003.

Timberlake, Craddock 2013: S. Timberlake, B. Craddock, “Prehistoric metal mining in Britain: the study of cobble stone mining tools based on artefact study, ethnography and experimentation”, Chungará (Arica) 45/1, 2013, pp. 33‑59.

Weller, Figuls, Grandia 2007: O. Weller, A. Figuls, F. Grandia, “Première carrière de sel gemme européenne: le Vall Salina à Cardona (Catalogne, Espagne) au Néolithique moyen (4500-3500 BC). Technologie, minéralogie et pétrographie de l’outillage lithique)”, in D. Morah, G. Dumitroaia, O. Weller, J. Chapman, L’Exploitation du sel à travers le temps, Piatra Neamt, Éditions Centre international de recherche de la culture Cucuteni/Institut d’archéologie Iasi, 2007, pp. 115‑133.

Zebrak 1995: P. Zebrak, “Mining hammers of Prljusa. Mali Sturac site, ancient mining and metallurgy in Southeast Europe”, Archaeological institute Museum of mining and metallurgy 199, 1995, pp. 37‑43.

Table des illustrations

Titre Fig. 1 – Map showing the location of Duzdağı in Nakhchivan (MBA).
Fichier image/jpeg, 1000k
Titre Fig. 2 – Duzdağı, survey, hammer types A and B (C. Hamon).
Fichier image/jpeg, 531k
Titre Fig. 3 – Duzdağı, survey, hammer types C and D (C. Hamon).
Fichier image/jpeg, 887k
Titre Fig. 4 – Duzdağı, survey, hammer type E (C. Hamon).
Fichier image/jpeg, 770k
Titre Fig. 5 – Duzdağı, survey, stone tools type F (C. Hamon).
Fichier image/jpeg, 724k
Titre Fig. 6 – Duzdağı, survey, diverse hammer types (C. Hamon).
Fichier image/jpeg, 833k
Titre Fig. 7 – Duzdağı, survey, grinding tools; a‑b: quern; c: grinder; d: pestles (C. Hamon).
Fichier image/jpeg, 710k
Titre Fig. 8 – Duzdağı survey, microphotographs of the thin section taken from a Type E hammer (no. 445) from Duzdağı; a‑b: phyric texture with large plagioclase crystals in a homogeneous recrystallized mesostase, scale bar 1 mm; c‑d: chlorite-rich mesostase with some epidote and calcite crystals, scale bar 200 µm; a and c under parallel nichols; b and d under crossed nichols (G. Fronteau).
Fichier image/jpeg, 2,2M
Titre Fig. 9 – Bronze Age mining cells in trench M1 (MBA, N. Gailhard).
Fichier image/jpeg, 845k
Titre Tab. 2 – Macro-lithic tools from the test trenches (2011‑2015).
Fichier image/jpeg, 83k
Titre Fig. 10 – Duzdağı excavations, Kura-Araxes trench M1; a: intermediate tool; b: pestle; c‑e: hammers; f: polisher (C. Hamon).
Fichier image/jpeg, 953k
Titre Tab. 3 – Hammer types from the test trenches (2011‑2015).
Fichier image/jpeg, 98k
Titre Fig. 11 – Iron Age slab‑extraction area from trench M1 (MBA, N. Gailhard).
Fichier image/jpeg, 829k
Titre Fig. 12 – Duzdağı excavations, Iron Age tools from M1; a: possible pick; b: intermediate tool; c‑d: pestles; e: grinding tool; f: massive hammer (C. Hamon).
Fichier image/jpeg, 631k
Titre Fig. 13 – Kültepe I, stone tools from the Kura-Araxes layers (after Abibullayev 1982); 6‑8 are possible hammers; 13‑14 are possible pestles (C. Hamon).
Fichier image/jpeg, 1,2M
Titre Fig. 14 – Kültepe I, hammer types belonging mainly to type E (collections of the National Museum of History in Baku, photos: C. Hamon).
Fichier image/jpeg, 631k



Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search