African Archaeological Review

, Volume 22, Issue 4, pp 177–197 | Cite as

Human Technical Behavior in the African Middle Stone Age: The Lithic Assemblage of Porc-Epic Cave (Dire Dawa, Ethiopia)

Article

The African origin of modern humans is the center of a large debate. Discoveries of anatomically modern human fossils in Sub-Saharan Africa correlated to lithic and faunal artifacts show that a “modern Behavior” is associated with the emergence of Homo sapiens. Even though the traits to define this modernity are sometimes difficult to apprehend, the study of the Middle Stone Age cultural phase is important for understanding the origin and evolution of the cognitive capacity of modern humans. Porc-Epic Cave in Ethiopia has a long sequence of Upper Pleistocene occupation. Several thousand bone and lithic artifacts were excavated during three major field excavations (1933, 1974, 1975–76). The lithic assemblage reveals that the relationship between humans and their environment is well organized and that the African terminology is sometimes difficult to apply. This paper proposes a synthesis of all the data, studies and conclusions I have made from the analysis of lithic materials from the 1933 and 1975–76 excavations in order to integrate Porc-Epic into the current debate of MSA and modern human Behavior.

L'origine de l'Homme moderne en Afrique fait l'objet actuellement d'un large débat. Les décourvertes de fossiles d'Homo sapiens en Afrique sub-saharienne associés à des industries lithiques montrent que, en parallèle à l'émergence de cette nouvelle espèce humaine, un comportement dit moderne s'est développé. Le Middle Stone Age est donc une période culturelle charnière dans l'évolution comportementale de l'homme. La grotte du Porc-Épic, en Éthiopie, est un témoignage de cette ère avec un remplissage pléistocène supérieur contenant des dizaines de milliers de restes fauniques et lithiques. L'étude de son important assemblage lithique révèle que le rapport de l'homme à la matière et à son environnement est très organisé et que le terminologie culturelle africaine est parfois difficile à utiliser.

KEY WORDS:

Prehistory horn of Africa Proc-Epic cave anatomically modern human middle stone age technical behavior 

INTRODUCTION

Recent studies reveal human Behavior changed significantly in the Middle Pleistocene (from 160 ka onwards and, more locally, from 300–230 ka). The studies also reveal that the Middle Stone Age (MSA) is closely related to the emergence of modern humans in Africa. This is shown by recent discoveries, as at Herto in Middle Awash (Ethiopia) where Homo sapiens idaltu fossils have been dated to ∼154–160 ka, and stratigraphically correlated with the Early Stone Age (ESA)/MSA transition (Clark et al., 2003; White et al., 2003). They also suggest that the appearance of Homo sapiens in Africa was anatomically and culturally gradual (Mellars, 1991; McBrearty and Brooks, 2000; Clark et al., 2003), and that it would have taken place in several areas at different times. The MSA shows traces of ‘cognitively modern’ Behavior, which is often visible in the technological, socio-economic and symbolic fields (Clark, 1989, 1993; Hayden, 1993; McBrearty, 1993; McBrearty and Brooks, 2000; Foley and Lahr, 2003; Henshilwood and Marean, 2003). At the typo-technology level, flakes, and tool types are more abundant and result from core reduction methods that become gradually more sophisticated. The Levallois, discoidal, and even blade technologies are among the most representative (technological modes 3 and 4). There is also a broader use of the natural resources, a tendency to specialize settlements, and more competent hunting. Lastly, the development of ornamentations, burials, and cave paintings represent novel symbolic traits. Some engraved or scraped pieces of ochre like those found in Blombos Cave in South Africa (Henshilwood et al., 2001) suggest a link between the level of symbolic activities and the evolution of modern Behavior (Henshilwood and Marean, 2003).

Considering that African MSA key sites are rare, the Porc-Epic Cave, in eastern Ethiopia, discussed here, with its thousands of faunal and lithic artifacts is particularly significant.

PORC-EPIC CAVE

Location and History of Research

Located at the southern gateway of the Afar Depression, 2 km south of Dire Dawa in Ethiopia (Fig. 1), the Porc-Epic Cave is in a cliff that rises abruptly above a river known as Datchatou. The cave was discovered at the end of 1920 by H. De Monfried and P. Teilhard de Chardin (Teilhard de Chardin, 1930). H. Breuil and P. Wernert from the French Muséum National d'Histoire Naturelle (MNHN) undertook the first excavation in 1933, mainly of a large part of the cave entrance (Breuil et al., 1951). Later, two excavation field seasons were led by J. Desmond Clark (1974) and K. D. Williamson (1975–76) from the University of California, Berkeley. A central trench of 6 m2 was dug in 1974. Most of the cave was then excavated in 1975–76 (Fig. 2). Lastly, in 1998, fieldwork was undertaken by a collaborative project between the Authority for Research and Conservation of Cultural Heritage (ARCCH) of Ethiopia and the French MNHN.
Fig. 1.

Geographical location of Porc-Epic Cave area (Dire Dawa, Harargue, Ethiopia).

Fig. 2.

Plan of the different excavated area and synthetic stratigraphic section N–S (transversal) of 9–10 W band (after Clark and Williamson 1984) of Porc Epic cave.

During successive excavations (1933, 1974, and 1975–76), a human half-mandible, showing both modern and archaic characteristics (Vallois, 1951), and tens of thousands of lithic and bone artifacts were recovered.

Furthermore, Assefa's study (2002) of a large part of the faunal remains will soon be published. It will provide important data on the occupation modalities: in particular, on the relationship between Porc-Epic occupants and their environment. The fauna shows a large variety of animal species and habitats. Taphonomic analysis reveals that humans were the primary agent for accessing faunal carcasses and that they were responsible for the accumulation of bones in the cave. During the supposed principal period of occupation (S.I.4), the possible contraction of the vegetation zones closer to the cave, suggested by the diversity of the species represented in the faunal assemblage and the geographical position of the site, would have facilitated access to resident and migratory faunal resources from diverse habitats. Small and agile mammals, like hares or hyraxes, which are difficult to catch, occur throughout the sequence of the cave. A number of them show cut marks, indicating that they contributed to the diet of the inhabitants, and that advanced tool kits were used for foraging and hunting purposes. This is an indication that the occupants of Porc-Epic were ‘cognitively' modern (Assefa, 2002).

Stratigraphy

The most complete stratigraphic section (Fig. 2) was exposed by Clark and Williamson (1984), and was completed during the 1998 mission. Correlation of the stratigraphic sections using a Harris Matrix (Fig. 3) helped in establishing some stratigraphic relationships. Some of the layers seen in different sections are identical.
Fig. 3.

Harris matrix showing the correspondences between different layers of Porc-Epic cave in different places studied during the 1998' fieldwork.

Four stratigraphic units, taking into account layers and field observations, were identified. These mark possible technological and socio-economical changes. Some refits suggest there were no strong post-depositional perturb-ations.

Dating

Several age measurements were undertaken after the 1974 excavation. The stalagmite unit that covered the breccia containing the principal MSA assemblage has been dated to 4590±60 and 6270±1020 years BP. These age estimates have to be taken cautiously because of post-depositional physical and chemical processes. According to Clark and Williamson (1984) they are not contemporary with the MSA occupation.

Nine samples of three MSA obsidian artifacts from the 1933 excavation stored at the Field Museum of Chicago were dated by obsidian hydration. The results range from 61202±958 to 71565±1575 years BP. Since hydration rate is influenced by temperature, and since the temperature then was probably lower than at present, dates obtained are minimal (Michels and Marean, 1984). Thus, the dating of the principal MSA occupation is uncertain.

PRESENTATION OF THE MATERIAL

Method of Study

This study is based on an analysis of part of the Porc-Epic lithic assemblage composed of more than 12000 artifacts. This work combined and compared the lithic material from the first 1933 excavation and those from more recent excavations in 1975–76 (Pleurdeau, 2003, 2004). It includes all the 1933 excavation collection (∼4000 artifacts) stored in Paris (Institut de Paléontologie Humaine) and the material of three squares of 1975–76 fieldwork (04N–05W; 06N–14W, 08N–07W) stored at the National Museum of Addis Ababa (∼8000 artifacts) (Fig. 2; Table I). The main objective was to reconstruct the chaînes opératoires by separating the various stages of production. However, since a large part of the material came from the 1933 excavation, for which only a few spatial data are available, the study of some distinctive spatial strategies was problematic. The research work currently conducted on all of the 1975–76 excavation material will be more significant as it represents several tens of thousands of artifacts from all the spatial and stratigraphical divisions of the cave.
Table I.

Typo-Technological Distribution and Spatial Origin of the Studied Material of Porc-Epic Cave (Dire Dawa, Ethiopia)

       

Total

Type of artifact

Zone

08N–07W square

05N–04W square

06N–14W square

1933 excavation

nb

%

Debitage products (>20 mm)

  

1432

661

1145

2321

5559

46

 

with:

Flakes

1109

429

795

1308

3641

30

  

Blades/small blades (bladelets)

298

193

309

824

1624

13

  

Points

 25

 39

 41

189

 294

 2

Debitage products (<20 mm)

  

431

234

1516

453

2634

22

Pebbles

  

  0

  0

  1

  4

   5

 0

Cores

  

 67

 42

 10

207

 326

 3

Waste

  

734

 28

1291

414

2467

20

Retouching products

  

 47

 14

108

  2

 171

 1

Retouched tools

  

 73

 83

 50

715

 921

 8

 

with:

Scrapers

 35

 22

 15

144

 216

 2

  

Denticulates

 10

  3

  2

 17

  32

 0

  

Becs

  6

  1

  1

 25

  33

 0

  

Notches

  1

  3

  0

 38

  42

 0

  

Perçoir

  2

  0

  0

  5

   7

 0

  

End-scrapers

  5

  1

  1

  8

  15

 0

  

Backed tools

  0

  1

  3

 25

  29

 0

  

Retouched points

 11

 36

 22

354

 423

 4

  

Composit tools

  3

 16

  6

 99

 124

 1

Total

  

2784

1062

4121

4116

12083

100

Acquisition of Raw Material

The inventory of raw materials (Fig. 4) used at Porc-Epic Cave shows that the main rock types exploited were flint (69%), basalt (17%), obsidian (8%) and sandstone/quartzite (3%).
Fig. 4.

Potential geological sources of raw materials used at Porc-Epic cave.

Ten different types of flint have been identified according to colors (white to orange and brown, or bicolor with marbling), textures (micrograined to grained), gloss (dull or glossy) and inclusions (microfossils, lithoclasts). These varieties of flints indicate geographic sources to be very close to the site. Indeed, the Jurassic carbonate sedimentary formations of the Somali plateau are very rich in flint nodules. During the 1998 field mission, it was observed that flint occurs near the cave along the Datchatou River.

Macroscopically, the basalt appears less varied than flint. Generally, the basalt is dense, with a very fine texture. Basalt is abundant in the depression, and, to a lesser extent, on the Somali plateau. However, easily accessible outcrops are rather rare.

Obsidian at the Porc-Epic Cave is varied. Black, brilliant, and translucent most of the time, it can also be dull and greenish. This greenish obsidian is not very common and is probably from the green obsidian flow of the Fantale Volcano, more than 200 km from the cave. Other possible sources of obsidian occur along the volcanic axial chain of Damahale and Gabilema in continuation of the Amiossa range. The few outcrops observed in the north of Dire Dawa were revealed by petrochemical analysis (Clark and Williamson, 1984). The closest and most probable sources are the volcanoes of Afdem and Assebot, located at about 130–150 km west of the cave.

Chemical analysis done by Agazi Negash and Steven Shackley (2006) confirm some of these sources. Some obsidian artifacts sampled from Porc-Epic assemblages and from several potential geological sources in the area have analogous chemical composition, and reveal that Ayelu, Assebot, and Kon'e are three identified geological sources, with a 150 km (Ayelu and Assebot) and 250 km (Kone) from the cave.

Quartzite, or rather a variety of sandstone/quartzite, have varied aspects. White to claret color, it may show marbled zonations. Like flint, this sedimentary rock appears to be local. Mesozoic sandy deposits crop out a few kilometres to SW of the cave.

Finally, sources of raw materials used at Porc-Epic Cave are different according to whether they are sedimentary or volcanic (see Fig. 4). Flints and sandstone-quartzite seem to come from local sources, whereas basalt and obsidian originate from further away with a minimal distance of more than 100 km from the cave for the latter. Nevertheless, the texture of the cortex on some flint artifacts suggests that flint could have been collected in the alluvial plains a few kilometres north of Dire Dawa. In addition, the rarity of primary flakes indicates that the first stage of knapping took place outside the studied zones and probably outside the cave. However, the large number of cores indicates that most of the core reduction took place at the site.

Production of Blanks

Four main processing schemes (schémas opératoires) occur throughout the entire sequence. Although the cores make possible an interpretation of the last steps of the reduction, the presence of some exclusive products make it possible to distinguish the different processes. Nevertheless, the schemes raise questions regarding blade production. Blades are numerous and their production is organized (edges and scars are rectilinear, and size is often standardized). Systematic study of the refits undertaken at the National Museum of Addis Ababa is particularly important for explaining the variation and sequence of the schemes (schémas opératoires) within one operative sequence (chaîne opératoire).

The first and second schemes are the most important (respectively 54 and 20% of the cores), involving a surface reduction as opposed to the more volumetric reduction of the fourth scheme (6%). The third scheme consists of an intermediate method between surface reduction and volume reduction (3%).

The blanks produced by these four schemes are diverse. Flakes are by far the most common (59%) of the debitage products. The frequency of blades (17%), bladelets (i.e., small blades with L < 35 mm) (13%) and points (7%) shows the range of the technical capabilities of the knappers at Porc-Epic Cave. All of these blanks are rather small in size, since the cores are, on average, less that 4 cm in size. Obsidian blanks, in particular, are much smaller.

In the first schéma opératoire (see Fig. 5), blank production results from a knapping axis parallel to the intersecting plane of the two core faces. This scheme is comparable, in many aspects, to the Levallois method. It allows the production of a wide range of products: flakes, blades, bladelets, and points. Cores (n=184) are rather typical. Often round, they show the opposition of two distinct surfaces. The first, when not completely obliterated by preferential removal, usually preserves radial negative scars that prepared the guide-ridges and convexity necessary for the extraction of the last flakes (or any type of product). The second surface, known as the “preparation surface of the striking platform,” is more convex and often highly prepared. Sometimes it shows negative invasive removals suggesting that this face was previously used as the production surface.
Fig. 5.

Stone artefacts knapped from 1st and 2nd schemas opératoires seen at Porc Epic cave: cores, 1 (preferential method), 2 (orthogonal recurrent method), 7 (“discoid” method), points, 3 (Levallois), 8 (pseudo-Levallois); blades, 4 (Levallois); flakes, 5 (orthogonal recurrent), 6 (unipolar recurrent).

Fig. 6.

Stone artefacts knapped from 3rd and 4th schemas opératoires seen at Porc Epic cave: cores, 1 (recurrent method), 2 (recurrent + preferential method), 3 (prismatic); blades, 4, 5; crested blades, 6, 7.

This scheme involves two distinct methods: one with a preferential flake (59%), and the other with recurrent flaking involving the extraction of a series of products in the same productive phase (41%). For both methods, backed products (débordants) were used for maintaining the optimal convexity of the surface in order to re-start the production.

The debitage products resulting from the first method are also typical. They are thin and mainly round, sometimes quadrangular. The majority often shows centripetal removals, indicating a high degree of preparation of the flaking surface. Butts are often faceted, thus confirming elaborate preparation of the striking platform.

Cores related to the second method show two distinct surfaces with a succession of unidirection, orthogonal or even radial removals, attesting to the recurrence of the process. In these cases, removals maintain the convexity and guide-ridges on the debitage surface necessary for the detachment of the next flake. In certain cases, convergent unidirectional removals seem to have allowed extraction of pointed flakes (n=96), known as Levallois points.

The second schéma opératoire (see Fig. 5) is less common than the first. It is characterized by one or two very convex flaking surfaces. Cores (n=62) always have two distinct surfaces and are exploited according to a non-parallel direction with the intersection plane between these two surfaces. The panel of debitage products seems less important than for the first scheme dominated by flakes. While the first scheme could be associated with Levallois debitage, this one is clearly connected to discoidal debitage, well known in Middle Stone Age and Middle Palaeolithic assemblages.

Cores are flaked following a radial pattern. A series of flakes is knapped recurrently around the core. Exploitation may concern one (40%) or both surfaces (60%), conferring a typical conical or biconical section to the core.

The debitage products are mainly flakes. Their scars and morphology (with a thick end-section) suggest a round-turning sequence of a sort of “uncapping” debitage. These flakes also tend to level the flaking surface. Therefore, the extraction of some products according to “cordale” direction, like pseudo-Levallois points, may have allowed the reduction of this flattening of the flaking surface and the restoration of a part of its convexity.

The third schéma opératoire (Fig. 6), with only nine cores, is quite rare in the assemblage, but occurs in several stratigraphic levels. The scheme appears as an intermediate form between the standard exploitation of a Levallois surface (First techno-type) and a standard volumetric blade reduction as in the Upper Palaeolithic (Fourth techno-type). The specific debitage products are rather difficult to recognize. Undoubtedly, the aim of the knapping process was to produce elongated artifacts, mainly blades and small blades (bladelets). Because it is difficult to recognize the methods employed through a simple reading of blade products, it remains difficult to recognize products resulting from this type of reduction. Distinct surfaces are still visible on the cores, one flake surface being transversely convex.

Just like the first scheme, two methods can be distinguished: one exclusively recurrent with a series of blades/bladelet production, and a second one which, while keeping the recurrent character, is marked by the final extraction of a large invasive removal.

The cores of the first method display uni- or bi-directional removal scars. The morphology of these cores is mostly semi-cylindrical suggesting that it could be a rough-out of typical blade debitage similar to standard European Upper Palaeolithic types.

The second method differs from the first one only by the final extraction of a blank that removes a large part of the surface. The unidirectional removals, knapped from an opposite striking platform, can be interpreted as having a double finality: production of bladelets and preparation of the knapped surface for the debitage of this last product. As all these cores come from the highest stratigraphic unit, the new age determinations currently in process will be of prime importance. Indeed, this sort of debitage looks like the Halfan method described in the Nile Valley sites like 1018 or E71P1C (Marks, 1968; Van Peer and Vermeersch, 1990; Wendorf and Schild, 1992). Morphological and technical resemblances also exist between this Porc-Epic method and the Levallois Halfan Method. The latter suggests a transitional techno-type between a surface debitage like the Levallois method and more volumetric debitage such as the Upper Palaeolithic bladelet reduction.

The last processing scheme (Fig. 6) involves a blade and bladelet production from cores that do not reveal any opposition between the two surfaces. It consists of a volumetric reduction of the core, in contrast to the surface reduction sensus stricto represented by the earlier schemes.

In this scheme, cores are few (n=21) but the quantity of blades and bladelets (n=1624) found in the assemblage, and the occurrence of crested blades (n=10) used for the initiation of this kind of debitage are significant, revealing this type of schéma opératoire. Except for the lower stratigraphic unit, each unit contains this type of core. They are mainly prismatic, with unidirectional and bi-directional scars showing the existence of a semi-revolving recurrent reduction on a large part of the core periphery. Some cores may also assume a more rectangular morphology while keeping the same technical characters.

Certain blade products, such as crested ones, seem to be exclusively specific to this scheme. Extracted from a projecting edge, these products create two parallel scars, initiating a repetitive process. The following blades and bladelets, not coming from the first reduction stage are abundant in the all sequence. Even if their technical process is difficult to recognize, they show by their morphology, often with parallel scars and edges, that a long sequence of blade debitage was a priority for the knapper. They are rather small, almost always less than 60 mm in length.

These reduction strategies appear to have a relative stability in their technical characteristics and numeric representation of the cores throughout the sequence (Table II). Thus no change or transition between MSA and LSA could be highlighted from a technical point of view.
Table II.

Technological Types and Stratigraphical Origin of the Studied Cores of Porc-Epic Cave (Dire Dawa, Ethiopia)

 

Stratigraphic unit

 

I (0–40 m)

II (40–60 cm)

III (60–200 cm)

IV (<200 cm)

Total

Type of core reduction

nb

%

nb

%

nb

%

nb

%

nb

%

First processing scheme:

          

One plane debitage surface cores

057

55

27

66

99

56

1

25

184

56

Second processing scheme:

          

One or two convex debitage surface cores

26

25

2

 5

32

18

2

50

62

19

Third processing scheme:

          

Intermediate reduction cores (surface/volume)

 5

 5

0

 0

 4

 2

0

 0

 9

 3

Fourth processing scheme:

          

Blades & bladelet tendance cores

 3

 3

5

12

13

 7

0

 0

21

 6

Other types

13

13

7

17

29

16

1

25

50

15

Total

104

100

41

100

177

100

4

100

326

100

The operational sequence (chaîne opératoire) of obsidian seems to be quite singular. Indeed, while it seems that there is no differential exploitation of the raw material for flake and blade production, the frequency of obsidian for bladelets/small blades (16%) is higher than that for any other type of blank. As a corollary, the frequency of bladelets among obsidian debitage products (29%) is higher than that for any other type of raw material (12% for flint) (Table III). It indicates that the exploitation of this non-local raw material took place on smaller nodules and was more intensive.
Table III.

Type of Debitage products (>20 mm) and Raw Material Distribution of the Studied Material of Porc-Epic Cave (Dire Dawa, Ethiopia)

 

Raw material

 

Flint

Basalt

Obsidian

Sandstone/Quartzite

Others

Total

Debitage product

nb

%

nb

%

nb

%

nb

%

nb

%

nb

%

Flakes

2601

65

637

68

192

54

125

75

86

75

3641

65

Blades

699

18

177

19

45

13

25

15

14

12

960

17

Small blades/bladelets

493

12

56

 6

104

29

 1

 1

10

 9

664

12

Points

197

 5

61

 7

16

 4

15

 9

 5

 4

294

 5

Total

3990

100

931

100

357

100

166

100

115

100

5559

100

Sequence of Tool Preparation

Tools are numerous at Porc-Epic Cave (see Fig. 7). Indeed, 17% of the debitage products are retouched. However, it seems that the rate of trimming differs according to the type of flakes. Points are the least numerous among debitage products, but they are proportionally the most retouched (26%) and thus retouched points are the most frequent tools. However, retouched points are made both on point blanks (48%) and flakes (35%). In general, there are two main categories of points: unifacial (75%) and bifacial (25%). In both cases, retouched points are small (over 90% being less than 60 mm in length) and elongated.
Fig. 7.

Panel of stone tools produced at Porc-Epic cave: unifacial points, 1 (with appointed base); bifacial points, 2 (mixes proximal retouches), 3 (large retouches); denticulates, 4; scrapers, 4–5 (lateral); backed microliths and bladelets, 7–9; end-scrapers, 10. Draws: M. Montesinos (1–6); D. Pleurdeau (7–9); after Breuil et al.1951 (10).

C. Perlès (1974) had defined some types of points according to the localization and the type of retouch and the final overall morphology of the artifacts. Her typology is used here with modification.

Unifacial points mainly belong to the triangular types with unilateral or bilateral retouch. Some of them show retouch on the proximal edge, producing an oval from. Certain points maintain or develop a morphological axis divergent from the technological axis of the initial blank. Moreover, some points are also pointed at the proximal part of the artifact.

Bifacial points, conversely, often show slimmer morphologies with a round retouched proximal edge in many cases. Sometimes, the lower face of the tool is completely covered by invasive retouch that may be thick or very fine, the latter suggesting the use of a soft hammer. It may also show only proximal retouch with a simple thinning of bulb and butt.

In all the stratigraphic units there is a wide typological range of scrapers. They are simple, double, or even triple, and are the most frequent tools after the points. One of the lateral edges, and sometimes both, is trimmed by direct, coarse retouch. For the majority of these scrapers, the blanks are flakes. End-scrapers are rare, very small and more finely worked. The end-scraper front is often round and flat, produced by small retouches. Notched tools (notches, denticulates, becs) are less frequent. They are often retouched on one of the lateral edges. They are made on thicker and larger flakes than scrapers.

Thirty-three backed bladelets or geometric tools are present. Such tools are characteristic of the African Late Stone Age and European Upper Palaeolithic and Mesolithic industries (Klein, 1995; Ambrose, 1998). At Porc-Epic Cave, these small tools were produced by backing the bladelets or flake segments. They occur in all the stratigraphic levels except in the lower unit IV. Half of them are obsidian. Backing is produced by short and abrupt retouch, conferring a slightly convex form to the edge. Sometimes it is also grossly curved, with ablation of the proximal and distal parts of the blank, producing a crescent.

DISCUSSION AND CONCLUSION

Porc-Epic Cave provides a perspective on the technical capability and Behavior of modern humans in East Africa during the Upper Pleistocene (Fig. 8). If the minimal dates of 70 ka for the occupation of some levels are confirmed by forthcoming determinations, Porc-Epic will be a key site for understanding the evolution of the human way of life in Ethiopia and in East Africa. In the entire sequence, without any significant typo-technological variations, the lithic assemblage reveals common Middle Stone Age features, both technological, such as Levallois and discoid core reduction methods, and typological, such as retouched points. It also presents some features that are commonly found in the Late Stone Age such as backed bladelets and geometrics. The occurrence of LSA elements in all the stratigraphic levels strongly indicates that both MSA and LSA elements co-existed right from the lower layers. In addition, the 1998 fieldwork in the cave revealed no evidence for strong bioturbation. This is confirmed by some refits of the 1975–76 excavation material, of artifacts from the same elevation.
Fig. 8.

Synthetic scheme of chaînes and schemas opératoires seen at Porc-Epic cave.

This association of MSA and LSA elements is rare in East Africa. Mumba industry from Mumba rockshelter in Tanzania has some similarities to the Porc-Epic assemblage. Even if there are some dating problems afflicting the site, Bed V and Bed IV, which have been dated by a range between 66 and 28 ka, include both MSA Levallois (debitage and retouched points) and LSA (geometric microliths) components. This industry was compared to the Howieson's Poort Industry (Mehlman, 1989, 1991), well known in South Africa (Klasies River, Boomblas or Border Cave), which also includes MSA and LSA elements (Deacon and Gelijnse, 1985; Deacon and Wurz, 1996; Singer and Wymer, 1982; Wurz, 1999).

The earliest complete LSA assemblage, Nasampolai industry, occurs at Enkapune Ya Muto rockshelter in Tanzania (also known as Twilight Cave), and is dated by several methods to around ∼50 ka. It consists almost exclusively of backed blades, bladelets and geometric microliths (Ambrose, 1998).

The present study suggests that the MSA and LSA cultural terminology is problematic and is perhaps in need of revision, as some authors have already proposed (McBrearty, 2000). If ancient local terminologies are now abandoned in favor of a general MSA and LSA terminology, it has often been difficult to interpret differences between assemblages because of the mixture of different levels (Brandt, 1986; Brandt and Gresham, 1989). Since Porc-Epic cannot by assigned exclusively to MSA, LSA or MSA/LSA transition, it may be that the transition from MSA to LSA occurred over a very long period of time (Clark et al., 2003; Tryon and McBrearty, 2002). In view of possible responses to local environment adaptation, these mosaic changes and transitions (both LSA/MSA and MSA/LSA) should be taken into consideration. Cultural units of the upper Middle Pleistocene and lower Upper Pleistocene might not have been simply a result of a general, gradual evolution, but also a product of local gradual evolution.

Lastly, the Halfan core reduction method seen in the Nile Valley sites (Marks, 1968; Van Peer and Vermeersch, 1990; Wendorf and Schild, 1992) is often described as an example of industry belonging to the Middle/Upper Palaeolithic transition, but the dates known so far are ∼17 to 26 ka range. It combines both Middle Palaeolithic (Levallois method) and Upper Palaeolithic (laminar method, backed tools) artifacts. The tendency is to attribute the origins of this transitional technology to the Middle Palaeolithic Group K (Van Peer and Vermeersch, 1990; Wendorf and Schild, 1992), which has recently been included into a “Lower Nile Valley Complex” marked by the absence of Nubian Levallois methods (Van Peer, 1998). This reduction method only occurs in the uppermost level at Porc-Epic Cave. Hence, the question arises: does the Porc-Epic assemblage present evidence for local evolution of debitage methods, or does it show a direct connection with the Halfan sites and the Nile Valley corridor? Is this method representative of LSA industry, MSA industry, or MSA/LSA transition? And does it mean that the upper stratigraphical unit of Porc-Epic dates to about this range (17—26 ka) or earlier, contemporary to the supposed origin of the method? The presence of microliths and Levallois and discoidal debitage throughout the sequence seems to negate the separation between MSA and LSA based solely on the presence/absence of these sorts of artifacts, as recent works reveal. Since the Nubian method (Guichard and Guichard, 1965; Van Peer, 1988) was discovered at K'one in Ethiopia (Kurashina, 1978), the Porc-Epic Halfan method is particularly important in supporting the role of East Africa and the Nile Valley in the technical and Behavioral diffusion hypothesis (Brandt, 1986; Clark, 1988; Vermeersch et al., 1990; Vermeersch, 1992, 2001; Van Peer, 1998; Kleindiest, 2000).

Notes

ACKNOWLEDGMENTS

Scientific and amicable help from several people has contributed to the realization of this study. So, I would like to thank Prof. H. de Lumley for allowing me to study the Porc-Epic lithic material stored in Paris; Sally McBrearty, M.-H. Moncel and P.-J. Texier for their advice; Dr. Yonas Beyene for helping and supporting me; Mrs. Mamitu Yilma, director of the National Museum of Ethiopia (NME) for her cooperation and for her acceptance for the study of the Porc-Epic material stored at the NME; and the Authority for Research and Conservation of Cultural Heritages (ARCCH). I would especially like to thank the Fyssen Foundation (Paris), which provided funding for my study at the NME (Addis Ababa). Many thanks also to the anonymous reviewers who gave me much advice and helped me to improve this paper.

REFERENCES CITED

  1. Ambrose, S. H. (1998). Chronology of the later stone age and food production in East Africa. Journal of Archaeological Science 25: 377–392.CrossRefGoogle Scholar
  2. Assefa, Z. (2002). Investigations of Faunal Remains from Porc-Epic: A Middle Stone Age Site in Southeastern Ethiopia. Ph.D. Thesis. State University of New York, Stony Brook.Google Scholar
  3. Brandt, S. A. (1986). The upper pleistocene and early holocene prehistory of the horn of Africa. The African Archaeological Review 4: 41–82.CrossRefGoogle Scholar
  4. Brandt, S. A., and Gresham, T. H. (1989). L'Âge de la Pierre en Somalie. L'Anthropologie 94: 459–482.Google Scholar
  5. Breuil, H., Teilhard de Chardin, P., and Wernert, P. (1951). Le Paléolithique du Harrar. L'Anthropologie 55: 219–228.Google Scholar
  6. Clark, J. D. (1988). The middle stone age of east africa and the beginnings of regional identity. Journal of World Prehistory 2: 235–303.CrossRefGoogle Scholar
  7. Clark, J. D. (1989). The origins and spread of modern humans: a broad perspective on the African evidence. In Mellars, P. A., and Stringer, C. B. (eds.), The Human Revolution. Behavioural and Biological Perspectives on the Origins of Modern Humans, Edinburgh University Press, Edinburgh, pp. 65–588.Google Scholar
  8. Clark, J. D. (1993). African and Asian perspectives on the origins of modern humans. In Stringer, C. B., Mellars, P. A. and Aitken, M. J. (eds.), The Origin of Modern Humans and the Impact of Chronometric Dating, Princeton University Press, Princeton, pp. 148–178.Google Scholar
  9. Clark, J. D., and Williamson, K. D. (1984). A middle stone age occupation site at Porc-Epic Cave, Dire Dawa (east-central Ethiopia), Part I. The African Archaeological Review 2: 37–64.CrossRefGoogle Scholar
  10. Clark, J. D., Beyene, Y., Woldegabriel, G., Hart, W. K., Renne, P. R., GILBERT, H., Defleur, A., Suwa, G., Katoh, K., Ludwig, K. R., Boisserie, J.-R., Asfaw, B., and White, T. D. (2003). Stratigraphic, chronological and behavioural contexts of Pleistocene Homo sapiens from Middle Awash, Ethiopia. Nature 423: 747–752.CrossRefGoogle Scholar
  11. Deacon, H. J., and Geleijnse, V. (1985). La Préhistoire d'Afrique du Sud: un aperçu. L'Anthropologie 89: 285–305.Google Scholar
  12. Deacon, H. J., and Wurz, S. (1996). Klasies River main site, cave 2: a Howiesons Poort occurrence. In Pwiti, G., and Soper, R. (eds.), Aspects of African Archaeology. Papers from the 10th Congress of the Pan-African Association for Prehistory and Related Studies, University of Zimbabwe Publications, Harare, pp. 213–218.Google Scholar
  13. Foley, R., and Lahr, M. M. (2003). On stony ground: Lithic technology, human evolution, and the emergence of culture. Evolutionary Anthropology 12: 109–122.CrossRefGoogle Scholar
  14. Guichard, J., and Guichard, G. (1965). The Early and Middle Palaeolithic of Nubia: A preliminary Report. In Wendorf, F. (eds.), Contributions to the Prehistory of Nubia, Forth Burgin Research Center and Southern Methodist University Press, Dallas, pp. 57–166.Google Scholar
  15. Henshilwood, C. S., Sealy, J. C., Yates, R., Cruz-Uribe, K., Goldberg, P., Grine, F. E., Klein, R. G., Poggenpoel, C., van Niekerk, K., and Watts, I. (2001). Blombos Cave, Southern Cape, South Africa: Preliminary Report on the 1992–1999. Excavations of the Middle Stone Age Levels. Journal of Archaeological Science 28: 421–448.CrossRefGoogle Scholar
  16. Henshilwood, C. S., and Marean, C. W. (2003). The origin of modern human behavior. Current Anthropology 44: 627–651.CrossRefGoogle Scholar
  17. Klein, R. G. (1995). Anatomy, behavior, and modern human origins. Journal of World Prehistory 9: 167–197.CrossRefGoogle Scholar
  18. Kleindienst, M. R. (2000). On the Nile corridor and the Out-of-Africa model. Current Anthropology 41: 107–109.CrossRefGoogle Scholar
  19. Kurashina, H. (1978). An examination of prehistoric lithic technology in east-central Ethiopia. Ph.D. Thesis. University of California, Berkeley.Google Scholar
  20. Hayden, B. (1993). The cultural capacity of Neandertals: A review and reevaluation. Journal of Human Evolution 24: 113–146.CrossRefGoogle Scholar
  21. Marks, A. (1968). The Halfan Industry. In Wendorf, F. (ed.) The prehistory of Nubia, Fort Burgwin Research Center, Texas, pp. 392–460.Google Scholar
  22. McBrearty, S. (1993). Reconstructing the environmental conditions surrounding the appearance of Modern Humans in East Africa. Culture and Environment: A Fragile Coexistence. Proceedings of the Twenty-Fourth Annual Conference of the Archaeological Association of the University of Calgary, pp.145–154.Google Scholar
  23. McBrearty, S. (2000). Identifying the Acheulian to Middle Stone Age transition in the Kapthurin Formation, Baringo, Kenya. Abstracts for the Paleoanthropology Society Meetings, pp. A12.Google Scholar
  24. McBrearty, S., and Brooks, A. S. (2000). The revolution that wasn't: A new interpretation of the origin of modern human behavior. Journal of Human Evolution 39: 453–563.CrossRefGoogle Scholar
  25. Mehlman, M. J. (1989). Late Quaternary Archaeological Sequences in Northern Tanzania. Ph.D. Thesis. University of Illinois, Urbana.Google Scholar
  26. Mehlman, M. J. (1991). Context for the emergence of modern man in Eastern Africa: Some new Tanzanian evidence. In Clark, J. D. (eds.), Cultural Beginnings. Approaches to Understanding Early Hominid Life-ways in the African Savanna. Römisch-Germanisches Zentralmuseum Mainz Forschungsinstitut für vor- und Frühgeschichte Monographie n. 19., R. Habelt, Bonn, pp. 159–176.Google Scholar
  27. Mellars, P. A. (1991). Cognitive changes and the emergence of modern humans in Europe. Cambridge Archaeological Journal 1: 763–776.Google Scholar
  28. Michels, J. W., and Marean, C. A. (1984). A middle stone age occupation site at Porc-Epic Cave, Dire Dawa (east-central Ethiopia), Part II. The African Archaeological Review 2: 64–71.Google Scholar
  29. Negash, A., and Shackley, M. S. (2006). Geochemical provenance of obsidian artefacts from the MSA site of Porc Epic, Ethiopia. Archaeometry 48(1): 1–12.Google Scholar
  30. Perlès, C. (1974). Réexamen typologique de l'industrie du Porc-Épic (Éthiopie): Les pointes et pièces pointues. L'Anthropologie 78: 529–552.Google Scholar
  31. Pleurdeau, D. (2003). Le Middle Stone Age de la grotte du Porc-Épic (Dire Dawa, Éthiopie): gestion des matières premières et comportements techniques. L'Anthropologie 107: 15–48.CrossRefGoogle Scholar
  32. Pleurdeau, D. (2004). Gestion des matières premières et comportements techniques dans le Middle Stone Age africain: les assemblages lithiques de la grotte du Porc-Épic (Dire Dawa, Éthiopie). British Archaeological Reports International Series 1317.Google Scholar
  33. Singer, R., and Wymer, J. (1982). The Middle Stone Age of Klasies River Mouth in South Africa, The University of Chicago Press, Chicago.Google Scholar
  34. Teilhard de Chardin, P. (1930). Le Paléolithique en Somalie française et en Abyssinie. L'Anthropologie 40: 331–334.Google Scholar
  35. Tryon, C. A., and McBrearty, S. (2002). Tephrostratigraphy and the Acheulian to Middle Stone Age transition in the Kapthurin Formation, Kenya. Journal of Human Evolution 42: 211–235.CrossRefGoogle Scholar
  36. Vallois, H. V. (1951). La Mandibule humaine fossile de la grotte du Porc-Épic près de Diré-Daoua (Abyssinie). L'Anthropologie 55: 231–238.Google Scholar
  37. Van Peer, P. (1988). La variabilité de la technologie Levallois dans la vallée du Nil égyptien. Revue archéologique de Picardie. Actes du colloque d'Amiens, 1986 “Culture et industries du Paléolithique en milieu loessique”. 12: 187–193.Google Scholar
  38. Van Peer, P. (1998). The Nile corridor and the out-of-Africa model. An examination of the archaeological Record. Current Anthropology 39: s115–s140.CrossRefGoogle Scholar
  39. Van Peer, P., and Vermeersch, P. M. (1990). Middle to Upper Palaeolithic transition: the evidence for the Nile Valley. In Mellars, P. A. (ed.), The Emergence of Modern Humans. An archaeological Perspestive. Edinburgh University Press, Edinburgh, pp. 139–159.Google Scholar
  40. Vermeersch, P. M. (1992). The Upper and Late Palaeolithic of Northern and Eastern Africa. In Klees, F., and Kuper, R. (eds.), New Lights on the Northeast African Past. Henrich-Barth Institut, Köln, pp. 99–154.Google Scholar
  41. Vermeersch, P. M. (2001). “Out of Africa” from an Egyptian point of view. Quaternary International 75: 103–112.CrossRefGoogle Scholar
  42. Vermeersch, P. M., Paulissen, E., and Van Peer, P. (1990). Le Paléolithique de la Vallée du Nil égyptien. L'Anthropologie 94: 435–458.Google Scholar
  43. Wendorf, F., and Schild, R. (1992). The Middle Paleolithic of North Africa: A status report. In Klees, F., and Kuper, R. (eds.), New Lights on the Northeast African Past. Henrich-Barth Institut, Köln, pp. 40–78.Google Scholar
  44. Wurz, S. (1999). The Howieson's Poort backed artefacts from Klasies River: An argument for symbolic behaviour. South African Archaeological Bulletin 54: 38–50.Google Scholar
  45. White, T. D., Asfaw, B., Degusta, D., Gilbert, H., Richards, G. D., Suwa, G., and Clark Howell, F. (2003). Pleistocene Homo sapiens from Middle Awash, Ethiopia. Nature 423: 742–747.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, Inc. 2006

Authors and Affiliations

  1. 1.Département de Préhistoire du Muséum National D'Histoire NaturelleInstitut de Paléontologie Humaine-1ParisFrance

Personalised recommendations