In this book I deal with a subject that has never been investigated before: dance at the beginning of agriculture. At first glimpse it seems that nothing can be said on such an elusive subject and that it lies beyond the boundaries of knowledgability. However, as we shall see below, the earliest art scenes in the ancient Near East and southeast Europe depict dancing. In the eighth to the fourth millennia BC this subject appears in many variations, covering a vast geographical expanse: the Levant, Mesopotamia, Iran, Anatolia, the Balkans, Greece, the Danube basin, and Egypt. There is plenty of evidence for this activity, almost four hundred depictions of dance are relevant to our study. Thus, dancing is the oldest and one of the most persistent themes in Near Eastern prehistoric art, and this theme spreads with agriculture into surrounding regions of Europe and Africa.

Dance, beside being a subject of enquiry in its own right, is used here also as a medium that sheds light on other interesting topics: the beginning of artistic scenes in the ancient Near East and southeast Europe, public calendrical rituals of early farmers, and various cognitive aspects concerning the dancing motif. The principal strategies used to promote the bonding of individuals into communities, and of individual households into villages, were public assemblies for the purpose of religious ceremonies. The archaeological examples discussed in this work are pictorial displays of this activity and shed light on it. The importance of these ceremonies is also borne out by ethnographic observations of modern pre-state communities, in which dance is indeed the most important component in religious ceremonies.

In periods before schools and writing, community rituals, symbolized by dance, were the basic mechanisms for conveying education and knowledge to the adult members of the community and from one generation to the next. The lengthy duration of dance depiction as a dominant artistic motif, together with its dispersion across broad geographical expanses (from west Pakistan to the Danube basin), testifies to the efficiency of the dancing motif as one of the most powerful symbols in the evolution of human societies.

Dance has been defined as a

complex form of communication that combines the visual, kinesthetic, and aesthetic aspects of human movement with (usually) the aural dimension of musical sounds and sometimes poetry. Dance is created out of culturally understood symbols within social and religious contexts, and it conveys information and meaning as ritual, ceremony, and entertainment. For dance to communicate, its audience must understand the cultural conventions that deal with human movement in time and space (Kaeppler 1992:196).

Dancing is an activity that is not limited to human behavior. As a means of communication, it has been observed in insects (the bee dance), birds, and various mammals' courtship interactions (von Frisch 1967; Wilson 1975:176-241, 314-335). However, as observed by McNeill (1995:13):

Community dancing occurs only among humans, if by that phrase we mean a form of group behavior whereby an indefinite number of individuals start to move their muscles rhythmically, establish a regular beat, and continue doing so for long enough to arouse euphoric excitement shared by all participants, and (more faintly) by onlookers as well. . . . Indeed, community dancing, together with marching and singing or shouting rhythmically is, like language, a capability that marks human off from all other forms of life. . . . Learning to move and give voice in this fashion, and the strengthened emotional bonds associated with that sort of behavior, were critical prerequisites for the emergence of humanity.

In human society, dance is a cross-cultural phenomenon that has been observed all over the world (Sachs 1952; Kraus 1969; Lange 1976; Bland 1976; Blacking and Kealinohomoku 1979; Clarke and Clement 1981; Cass 1993; S. J. Cohen 1998). It has been suggested that dance, as a medium of nonverbal communication, was already practiced during the Paleolithic era (Louis 1955; Blacking 1976; McNeill 1995:13-35). Pictorial sources such as rock art and portable items display dancing in past nonliterate societies. The earliest examples of these have been reported from Paleolithic European art, such the cave at Cala dei Genovesi on the island of Levanzo near Sicily and in Addaura Cave, near Palermo in Sicily (Leroi-Gourhan 1967:381-382, Fig. 710; Holloway 1991:2-4, Figs. 4-5). Archaic rock-art depictions of dance, whose dating is not always clear, have been reported from various parts of the world, such as Italy, Turkey, Israel, Azerbaijan, and India (Anati 1955, 1964, 1994; Peschlow-Bindokat 1995; Dzhafarzade 1973; Brooks and Wakankar 1976:18-19; Malaiya 1989, 1992; Neumayer 1997). More recently, rock art depicting dancing scenes has been produced by Australian aboriginals (Godden 1982, Figs. 25-27; Walsh 1988, Figs. 83, 111) and San Bushmen of Southern Africa (Vinnicombe 1976:307-319; Lewis-Williams and Dowson 1989, Figs. 14-18; Lewis-Williams 1999). Historical texts and depictions inform us about dancing in the great civilizations of Mesopotamia (Matou_ová 1970, 1993; Collon 1987:151-153), Egypt (Lexová 1935; Brunner-Traut 1958, 1985; Wild 1963; Baldacci 1987; Saleh 1998), Biblical Israel (Gruber 1981; Mulder 1992), early Aegean (Evans 1930:66-80; Lawler 1964:40-57; Iakovidis 1966; Dothan 1982:237-249; Goodison 1989; Lefèvre-Novaro 2001), classical Greece (Lawler 1964; Prudhommeau 1965), Rome (Kraus 1969:40-45), and the Middle Ages (Molé 1963; Kraus 1969:46-62). Dancing appears now in every form of human organization: urban, rural, pastoral, or hunter-gatherer communities (S. J. Cohen 1998; Reed 1998).

In this study I will limit discussion to the evidence for dancing activities in a specific chronological, geographical, and socioeconomic milieu, that is, the village communities of the Near East and southeast Europe, from ca. the eighth to the fourth millennia BC (calibrated). All the dates used in this work are calibrated BC, based on 14C radiometric datings.

The term "village communities" refers to what is commonly called the Natufian, Neolithic, and Chalcolithic periods. At this stage of human history the Paleolithic way of life based on small bands of nomadic hunter-gatherers disappeared, while cities and states, which emerged during the second half of the fourth millennium BC in Mesopotamia, had not yet developed. This period is the subject of many studies, both of the Near East and of southeast Europe (see, for example, Braidwood and Braidwood 1953; Mellaart 1975; Redman 1978; Nissen 1988; Bar-Yosef and Belfer-Cohen 1989; Maisels 1991, 1993; Tringham 1971; Hodder 1990). The village communities lasted some nine thousand years (from ca. 12,000 to 3,000 BC) in the Near East, and during this time, rapid changes took place in almost every area of human existence: Permanent village-type settlements developed (Flannery 1972a). Population growth and agglomeration of large communities took place (M. N. Cohen 1975; Hassan 1981). Food production was initiated, with an increasing number of domesticated plants and animals (Bender 1975; Zohary and Hopf 1988; Flannery 1973; Davis 1987:126-154; Buitenhuis and Clason 1993). Technological innovations, which included an increased use of pyrotechnology, were developed, initially for the production of plaster (Gourding and Kingery 1975; Kingery et al. 1988) and later for pottery and metals as well. Social developments produced stratification and the emergence of political institutions (Flannery 1972b; Redman 1978). Finally, mythology and basic concepts of religion were established (Amiran 1962; Margalit 1983; Garfinkel 1994).

Material culture and subsistence are usually easier to examine and analyze than social organization, religion, and ideology. The dancing scenes discovered in the early village communities are related to the second group and thus shed light on issues that are difficult to investigate.

The Beginning of Artistic Scenes in the Ancient Near East

Using the term "art" nowadays seems to need semantic, as well as conceptual, clarification. There is a tendency by some scholars to reject the term when dealing with prehistoric images and to introduce other phrases, including "symbolic expressions," "imagery," and the like. Soffer and Conkey wrote, "Long-standing debates about the definition(s) of 'art' conclude that aesthetic function is something that we cannot assume to have been the case in Prehistory. In fact, ethnographic data from nonwestern cultures clearly show us otherwise" (1997). There is a disturbing point regarding the rhetoric of that statement, as if the opinion of the authors should be considered as a "conclusion." Indeed, this ethnocentric, narrow-minded approach, as if only the western civilization has "art" while other human societies have "imagery," is rightly rejected by many archaeologists and anthropologists. As summarized by Morphy (1989:1):

Researchers have tried to solve the problem by changing the name; by switching from "art object" to "image" or "representation" or "information system" on the grounds that these are more neutral, less value-laden terms. Yet the replacement terms do not really help the situation. They are often more narrow in their definition than "art" itself which, because of all the argument over what it is, can have the advantage of being broadly conceived, whereas "representation", for example, may apply to only one aspect of an object.

The same approach has been expressed by ethnographers: "Art from non-Western cultures is not essentially different from our own, in that it is produced by individual, talented, imaginative artists, who ought to be accorded the same degree of recognition as western artists" (Price 1989). On the matter of aesthetic, Gell specifically noted:

There is no sense in developing one "theory of art" for our own art, and another, distinctively different theory, for the art of those cultures who happened, once upon a time, to fall under the sway of colonialism. If Western (aesthetic) theories of art apply to "our" art, then they apply to everybody's art, and should be so applied (Gell 1998:1).

Language is a mean of communication, but it can also become a means of miscommunication. The developing of a highly specialized jargon in a discipline may cause a better understanding among the members of the discipline, but it will be a barrier to the people outside. "Art," however, term is understandable to every layperson. It is an easy communication devise between the archaeologist and the general public. By using terms such as "imagery," we immediately disconnect ourselves from the general public. Indeed, even the above-cited words of Conkey and Soffer were published in a book ingeniously called Beyond Art: Pleistocene Image and Symbol. In this way, the rejected term "art" was incorporated into the book's title.

After clarifying the legitimacy of the term "art," the context of the depictions of dance within the development of art history should be examined. The nomadic hunter-gatherer societies of the Upper Paleolithic and Epi-Paleolithic in the Near East hardly produced artistic expressions at all, and only isolated art objects have been reported from this region. With the establishment of sedentary settlements in the ancient Near East, this situation altered completely. The earliest permanent villages were built in the Natufian culture of the Levant, dated to as early as the twelfth millennium BC. Natufian sites are characterized by rich symbolic artwork, falling into the following groups:

• Anthropomorphic Figures. Small figurines made of bone or stone have been reported from el-Wad Cave and Eynan (Perrot 1979, Fig. 17; 1966, Fig. 23:1-2).
• Zoomorphic Figures. Small figures made of bone or stone, sometimes forming the decorated handle of a larger tool, have been reported from el-Wad and Kebara Caves, Nahal Oren, and Wadi Hammeh (Perrot 1979, Figs. 10, 12, 15; Noy 1991, Figs. 5, 6:1-4; Stekelis and Yizraely 1963, Pl. 4:A-D; Edwards 1991, Fig. 9:2).
• Geometric Engravings. Geometric patterns, including meanders, were engraved on various stone vessels, stone slabs, or bone tools. These have been reported from Eynan, Nahal Oren, Hayonim Cave, Wadi Hammeh, and Shukba Cave (Perrot 1966:467, Figs. 15:9-10, 23:4; Noy 1991, Figs. 2:5-6, 3, 4:1; Belfer-Cohen 1991, Fig. 3:6; Edwards 1991, Figs. 6:7-8, 8, 10).
• Body Ornaments. Large quantities of necklaces, bracelets, and belts, made of dentalium shells and various animal bones or teeth, have been unearthed in graves, in situ on the skeletons. Examples of these have been reported from el-Wad Cave, Eynan, and Hayonim Cave (Garrod and Bates 1933, Pls. VI:1-2, VII; Perrot and Ladiray 1988, Figs. 14, 15, 18, 22; Belfer-Cohen 1988:302).
• Varia. One item made of a river pebble depicts a couple in the act of sexual intercourse. However, Boyd and Cook (1993) questioned the object's Natufian context. This figurine, if indeed Natufian, is the earliest depiction of a scene in the ancient Near East. In any case, since no similar object has been discovered anywhere, it is an isolated phenomenon without any influence on Natufian or Neolithic art.

In the early ninth millennium BC, the period commonly designated Pre-Pottery Neolithic A (PPNA), the corpus of symbolic depictions unearthed includes the following groups:

• Anthropomorphic Figurines. Figurines of clay or stone have been reported from Gilgal I, Netiv Hagdud, Salabiya IX, and Nahal Oren (Noy 1989, Fig. 5:2-5; Bar-Yosef 1980; Bar-Yosef et al. 1991, Fig. 13; Stekelis and Yizraely 1963, Pl. 2:F-H).
• Zoomorphic Figurines. At Pre-Pottery Neolithic A sites the zoomorphic figures are birds: Mureybet Layer III, Gilgal I, Nemrik 9, and Hallan Çemi Tepesi (Pichon 1985; Noy 1989, Fig. 5:1; Kozlowski 1997; Rosenberg 1994, Fig. 12:1). No definite cattle figurines have been reported from this period.
• Incised Stone Bowls. A zoomorphic figure and geometric patterns engraved on stone bowls have been reported from Hallan Çemi Tepesi (Rosenberg and Davis 1992, Fig. 8:1-13).
• Incised Stone Tools. Geometric and meanderlike patterns were engraved on various ground stone tools. Meander patterns have been reported from Netiv Hagdud, Mureybet, and Jerf al Ahmar and a net pattern from Gilgal I (Bar-Yosef et al. 1991, Fig. 12; Cauvin 1985, Fig. 2:1; Stordeur 1998; Noy 1989, Fig. 4:1). These motifs seem to continue the Natufian tradition described above. In addition, an item decorated with zoomorphic figures has been reported recently from Jerf al Ahmar (Stordeur 1998, Fig. 9:1).
• Architectural Decoration. A totally new area of symbolic expression in the beginning of the Neolithic period has recently come to light. In two sites of the northern Levant, excavators unearthed decoration from buildings. In Göbekli Tepe the decoration includes incisions of zoomorphic figures and geometric patterns. These were executed on large limestone T-shape pillars (Beile-Bohn et al. 1988; K. Schmidt 1998). In Jerf al Ahmar a mudbrick bench was decorated with a geometric pattern in relief (Stordeur et al. 2001, Fig. 8).

In the next chronological stage, the late ninth and the eighth millennia BC, the period commonly designated Pre-Pottery Neolithic B (PPNB), there is a rapid growth of symbolic expression. Examples now include the following:

• Anthropomorphic Figurines. These are reported from sites all over the Near East and usually include female figures and much less frequently also male figures: Jericho, Beidha, Nahal Hemar Cave, 'Ain Ghazal, Munhata, Tell Aswad, Cafer Höyük, Nevali Çori, Jarmo, Sarab, and Çayönü (Kenyon and Holland 1982, Fig. 224:2; Kirkbride 1966, Fig. 4:1; Bar-Yosef and Alon 1988, Pl. IX; Rollefson 1983; Garfinkel 1995, Figs. 13-14; de Contenson 1983, Fig. 7; Cauvin 1989, Fig. 11; Hauptmann 1991-2:32, Fig. 27; Broman Morales 1983, 1990).
• Zoomorphic Figurines. At this stage, cattle make their first appearance, replacing the bird figurines of the previous period. They have been discovered in sites all over the Near East: Jericho, 'Ain Ghazal, Abu Ghosh, Munhata, Tell Ramad, Jarmo, Sarab, and Çayönü (Kenyon and Holland 1982, Fig. 224:6-9; Rollefson 1983; Lechevallier 1978, Fig. 35:1-2; Garfinkel 1995; de Contenson 1981; Broman Morales 1983, 1990). These figures are so schematic that it is not always clear if they indeed represent cattle or rather other four-legged animals such as sheep or dogs. Other types of zoomorphic figures were also found, including an ibex at Beidha, a rodent at Nahal Hemar Cave, and pigs at Munhata, Jarmo, and Sarab (Kirkbride 1966, Fig. 4:2; Bar-Yosef and Alon 1988, Pl. XII:1; Garfinkel 1995, Fig. 17:5; Broman Morales 1983, Fig. 155:6, 1990, Pl. 2).
• Engraved Slabs. Small stone slabs decorated with incisions portraying various animals have been reported from the desert campsite of Dhuweila (Betts 1987).
• Remodeled Skulls. Human skulls plastered or covered with bitumen have been reported from the Levant: Jericho, Tell Ramad, Beisamoun, 'Ain Ghazal, Nahal Hemar Cave, and Kfar Hahoresh (Kenyon 1957, Figs. 19-20, 22; 1981, Pls. VIII:b-d, IX; de Contenson 1967:20-21; Lechevallier 1978:150; Rollefson 1983; Bar-Yosef and Alon 1988, Pl. XXIII; Goring-Morris 2000, Fig. 3).
• Anthropomorphic Statues. Two different groups can be subsumed under this category: statues made of clay and of plaster, 40-90 cm in height, discovered at Jericho, Tell Ramad, 'Ain Ghazal, and Nahal Hemar Cave (Garstang et al. 1935:166; de Contenson 1967:20-21; Rollefson 1983; Bar-Yosef and Alon 1988, Pl. VIII); statues made of elongated limestone monoliths, 40-100 cm in height, were discovered at Nevali Çori in the northern Levant (Hauptmann 1993).
• Masks. Life-sized limestone masks have been excavated in the southern Levant: Nahal Hemar Cave and Basta. A few such masks with no clear archaeological contexts have also been reported (Bar-Yosef and Alon 1988, Pl. X; Nissen et al. 1987, Fig. 16:1; Perrot 1979, Figs. 20, 26; Noy 1999:123-133).
• Decoration of Architectural Components. Three different groups can be distinguished here: painted plaster floors, discovered at 'Ain Ghazal and Tell Halula (Rollefson 1990; Molist 1998); painted walls of Bouqras (Akkermans et al. 1983); and monumental stone pillars engraved with anthropomorphic figures in a cultic structure at Nevali Çori (Hauptmann 1993).
• Dancing Scenes. The first appearance of scenes in the art of the ancient Near East took place at this stage. These have been reported from three Pre-Pottery Neolithic B sites in the Levant: Nevali Çori, Tell Halula and Dhuwila (Figs. 7.3:a, 7.4, 7.6:a). All three have a common subject--they depict dancing figures.

Rollefson suggested subdividing the art objects from 'Ain Ghazal into two groups based on their size (1983, 1986). This idea can be applied to objects from the Pre-Pottery Neolithic period in general. In the first group there are small anthropomorphic and zoomorphic figurines, usually smaller than 10 cm. These were used by individuals and reflect household cults. In the second group there are the large anthropomorphic statues, plastered skulls, and masks, usually larger than 20 cm. These were used on more public occasions and reflect rituals performed on the community level. The objects of both groups, however, are isolated art and cult items. Even when found in large numbers together, such as the plastered skulls or the anthropomorphic statues, they do not represent any scene, and no interdependence between the objects in any given concentration can be observed.

Following Rollefson and adding to his original observations, it seems that a three-stage development can be pointed out. First, small art objects were manufactured and these were used on the family level. Second, larger art objects were produced, and these were used on the community level. The third stage is marked by the appearance of the first scenes. This is a most dramatic change as the scenes described a few people acting together. They have the capacity of storing and transmitting much more information than any single art object. Thus, the scenes represent a big step towards writing. The importance of scenes was summarized by Renfrew and Bahn (1991:368):

Painting, drawing, or carving on a flat surface in order to represent the world offers much more scope than the representation in three dimensions of a single figure. For it offers the possibility of showing relationships between symbols, between objects in the cognitive map. In the first place, this allows us to investigate how the artist conceived of space itself, as well as the way in which events at different times might be shown. It also allows analysis of the manner or style in which the artist depicted the animals, humans, and other aspects of the real world.

Scenes, as opposed to figurines, are rather rare in the protohistoric Near East and southeast Europe. They can be classified in three main categories:

• Wall Painting. The most well known of these are the Çatal Höyük wall paintings and reliefs, a unique assemblage (see, for example, Mellaart 1967; de Jesus 1985; Hodder 1987; Forest 1993). Another example of a wall painting depicting a scene has been reported from Umm Dabaghiyah (Kirkbride 1975, Fig. 7a).
• Scenes Displaying Dancing Figures. This is the largest category in terms of chronological duration, geographical distribution, and the number of examples reported (Garfinkel 1998). It is therefore rather surprising to see that these dancing figures have never been the subject of a comprehensive analysis, and only some aspects of them have been discussed (Herzfeld 1941:29-42; Mesnil du Buisson 1948:23; Parrot 1960:44-46; Gulder 1960-1962; Vanden Berghe 1968; Nitu 1970; Marinescu-Bîlcu 1974a; Meyerhof and Mozel 1981; Yakar 1991:314; Esin 1993; Mantu 1993; Matou\xová 1970, 1993).
• Nondancing Scenes. These are known on pottery vessels depicting human figures together with animals or architectural elements. This rare category has been reported from only a few sites. The most interesting examples come from Tell Halaf, Arpachiyah, Tepe Gawra, and a vessel of unknown provenance from Iran (von Oppenheim 1943, Fig. LX; Hijara 1978; Ippolitoni-Strika 1990; Breniquet 1992a; Tobler 1950, Pl. LXXVIII:a-b; Amiet 1979, Fig. 15-17).

This study is devoted to the second category, the dancing scenes. The three main subjects to be dealt with here are the dancing performance (Chapter 2), the social context of the dance (Chapter 3), and cognitive aspects of the dancing scenes (Chapter 4).

Research Objectives

Nine different research objectives, all related to the subject of dance and dancing scenes, are presented below. Although I may not address all the relevant questions, nor provide all the possible answers, I hope that drawing attention to this neglected topic will provoke further inquiry and discussion.

1. Relevant data

The Neolithic and Chalcolithic dancing scenes from the Near East and southeast Europe have never been previously collected or presented together. The evidence is dispersed over hundreds of publications, many of them obscure excavation reports. Therefore, the first aim of this work is to acquaint the reader with the different sites and the relevant finds unearthed in each of them. Most of the items are also presented here in drawings. In some cases the text refers to additional identical items that were discovered at the same site in large quantities and are thus not included in the illustrations. All this data is presented in Part II of this work.

The distribution of the motif dictates the chronological and geographical boundaries of my study. Chronologically the earliest dancing scenes appeared in the eighth millennium BC, and they continue to dominate the artistic record, at least in one region--the Levant--until the third millennium BC. The geographical range of the phenomenon includes the Levant, Mesopotamia, Iran, western Pakistan, Anatolia, the Balkans, Greece, the Danube basin of southeast Europe, and Egypt. The motif does not appear in all the areas at the same time. Thus the presentation of the data follows six major chronological-geographical units: Neolithic Near East; Halafian and Samarra cultures; Neolithic and Chalcolithic Iran; Neolithic southeast Europe; Predynastic Egypt; and later examples from the Near East. All together, some 170 sites and nearly 400 items are included. In an appendix, other possible examples of dancing figures in the Neolithic Near East are examined in the light of previous reconstructions. The dancing scenes, it should be emphasized, were a most popular, indeed almost the only subject used to describe interaction between people in the Neolithic and Chalcolithic periods.

The present study of dancing scenes is derived from my general understanding of the current state of art historical research. This research in ancient Europe and the Near East has usually been classified according to the following fields of specialization: on the one hand, the art of Paleolithic Europe, which represents the symbolic expressions of hunters and gatherers; on the other hand, the various art traditions of Mesopotamia, Egypt, Greece, and Rome, each type representing the symbolic expressions of states and empires. Art historical research on early village communities in the Neolithic and Chalcolithic periods, which lies between the two fields mentioned above, has not been extensively developed. Only very seldomly has a specific monograph on the art of the Neolithic or Chalcolithic periods been published (Goff 1963a; Cauvin 1972, 1994; Dumitrescu 1974; Gimbutas 1982, 1989; von Wickede 1990).

The leading authorities on ancient Near Eastern art have not paid much attention to the Neolithic period. The concepts that prevailed in the 1950s and 1960s concerning the art and cult of this period can be found in the works of eminent authorities. Henry Frankfort wrote: "The prehistoric clay figurines of men and animals do not differ in character from similar artless objects found throughout Asia and Europe. A history of art may ignore them, since they cannot be considered the ancestors of Sumerian sculpture" (Frankfort 1955:2). Edith Porada wrote: "For this early period [the eighth millennium BC] we cannot assume the existence of concepts of anthropomorphic deities similar to those later known in the cultures of the ancient Near East" (Porada 1965:21). In various books and catalogues on the art of the ancient Near East, the Neolithic and Chalcolithic periods usually occupy a brief chapter between the introduction and Sumerian art.

However, over the years, Neolithic art objects from various excavations have gradually accumulated. Nowadays there is no justification for an approach that ignores the art of the Neolithic and Chalcolithic periods, since in the last decade, archaeologists have unearthed and published art objects from dozens of Neolithic sites, of previously unknown quality and quantity. Thus, another objective of this research is to draw scholarly attention to the possibility that art historical research of the Neolithic and Chalcolithic periods can now be developed into an independent field of art history, that of early village communities.

2. A methodological framework for the depiction of dance

To analyze displays of dance, a methodological framework is required. Since, to the best of my knowledge, this topic has not been subject to systematic methodological consideration, it is necessary to propose one here. The next section of this chapter raises and discusses thirteen methodological aspects of the depiction of dance.

3. Structural analysis of dance

Analyzing the form and style of the performance carried out in the early village communities contributes to the field of dance history, which, when dealing with the evidence from antiquity, usually concentrates on Greek vases. To obtain the maximum amount of information, I have defined detailed categories of analysis, and the different items are analyzed in accordance with these categories. Not all the scenes previously defined as dance fit these categories, and thus the interpretation of some wall paintings from Çatal Höyük as a "dance of the hunters" is rejected below.

4. Emphasis on the cultic nature of dancing

The study of the cult and religion of ancient societies has lately attracted considerable attention. The dancing scenes can contribute to our understanding of public religious ceremonies in the early village communities in the Near East and southeast Europe. This matter also takes into consideration linguistic aspects and ethnographic observations.

5. Clarification of linguistic aspects associated with dance

It has been previously noted that dancing terms in West Semitic languages are loaded with more than one meaning (Mandelkern 1896:369; Loewenstamm 1965; Wensinck 1986). The dancing scenes add a new dimension to the linguistic context of the following terms: dance, festival, going in a circle, mourning, and pilgrimage.

6. Functional analysis of dance

The dancing scenes should be understood against the background of their social context, that is, the village communities of the Neolithic and Chalcolithic periods. Ethnographic observations are used to gain a wider view of dancing in pre-state societies. Many observations by modern anthropologists of traditional societies are devoted to dance. There is a vast literature on specific case studies, as well as works devoted to the anthropology of dance within a wider theoretical framework for investigating the significance of dance (see, for example, Rust 1969; Lange 1976; Royce 1977; McNeill 1995). As dancing is a universal phenomenon, and as the significance of dance has been investigated by dance scholars, the basic question discussed here is not why people dance but rather why dancing was used as a motif in the art of early village communities. In other words, it seems unlikely that the people in the early village communities were dancing many more hours per week than hunter-gatherers or city dwellers of the ancient Near East, even though they emphasized dancing activities in their symbolic expressions.

Of special interest to my study are observations on the San Bushmen of Southern Africa, since there exists abundant documentation of both the dancing activities and artistic symbolic expressions of this people. In Bushmen societies, dance activities are extremely important in daily life (Marshall 1969; Katz 1982; Biesele 1978). Nevertheless, dance is not a major motif at all in artistic expressions (Vinnicombe 1976:307-319; Lewis-Williams 1981:19; Lewis-Williams and Dowson 1989; but see Lewis-Williams 1999; Garfinkel 1999b). This example demonstrates that there is no direct correlation between the role of dance in daily life and the appearance of dance in symbolic expressions. Two questions are relevant here: (1) why were dancing scenes the most popular, indeed almost the only subject used to describe interaction between people in the Neolithic and Chalcolithic periods? and (2) why do dancing scenes lose their prominent position with the rise of urban societies in the Near East? To answer these questions, Chapter 3 deals with social developments in the eighth to fourth millennia BC in the Near East and southeast Europe.

7. Replacement of the dancing motif

The objective of this section is to identify a widely distributed motif in the ancient urban Near East, which expresses social interaction between people and thus replaces the dance motif.

8. Cognitive aspects of the dancing scenes

In a programmatic statement in the introduction to the first volume of the Cambridge Archaeological Journal, Scarre declared that "human cognition is one of the key areas of archaeological thinking at the present day, encapsulating a wide range of those activities which give the human species its unique character and separate us most clearly from our near relatives the non-human primates: art, symbolism, mensuration, religion and language" (Scarre 1991). As dancing activity is closely associated with public religious ceremonies, dancing opens a window onto the cognitive map of the early village communities' rituals.

There is a growing archaeological literature on issues such as ritual and cultic performance. These works deal mainly with the analysis of cult objects and temples, that is, the material remains of ritual activities. Marcus and Flannery, for example, combined three methodological approaches when dealing with ancient Zapotec ritual and religion: (1) the direct historical approach, (2) the analysis of public space and religious architecture, and (3) the contextual analysis of religious paraphernalia (Marcus and Flannery 1994:55). The nature of the evidence concerning dancing, however, is totally different from the source material of Marcus and Flannery, since we do not have direct remains of the activity. "Direct remains" of dance in the archaeological record could theoretically be footprints or an arrangement of skeletons trapped during dance activity. The suggestion that circular platforms of carefully constructed ashlar masonry stones discovered at Knossos were used for circle dances (Warren 1984) is problematic, since the platforms may have been used for other purposes. What we do have in hand from the past are depictions of dance that are not "objective" photographs but the results of cognitive processes that function as filters, eliminating the less important aspects of an event and emphasizing its essence. Depictions of dance may give rise to all kinds of theoretical or methodological issues, since they are not direct material remains of performance and ritual. This is where cognitive analysis comes in, focusing on the process that transfers reality into depiction. The importance of pictorial relationships has been summarized by Renfrew and Bahn as follows (1991:363):

We can obtain the greatest insight into the cognitive map of an individual or a community by representation in material form of that map, or at least a part of it. . . . But the more general case is that of depiction, where the world, or some aspect of it, is represented so that it appears to the seeing eye very much as it is conceived in the "mind's eye."

Dancing scenes are clearly suitable for the application of the approach labeled "Cognitive-Processual Archaeology" (Renfrew and Bahn 1991:431-432, and see discussion there), as the subject of dance implies the following points:

1. As ethnographers have noted, dance is a cross-cultural phenomenon in human behavior and has been observed the world over. It appears in every form of social organization: urban, rural, pastoral, and hunter-gatherer.
2. From a structural point of view there are biological limits to the movements of the human body, which dictate possible body gestures (Laban 1971; Fitt 1988).
3. From a functional point of view there are similarities in the way traditional human societies use dance: it is associated with public religious activities.

9. Communicative aspects of the dancing scenes

This research objective concerns not the dancing activity but the function of the objects on which the dancing scenes were depicted. Pottery vessels, which are the most common objects decorated with dancing scenes, are examined as a means of information exchange.

Methodological Remarks on the Display of Dance

Prehistoric displays of dance raise all kinds of methodological problems, some of which are specifically derived from the subject of dance, while others are related to a more general level of analyzing archaeological material. Identifying a scene as dance activity is not always self-evident. Some items that are interpreted here as dancing scenes have previously been understood differently. For example:

1. The depiction presented in Figure 7.6:b has been described as "a couple of deities...shown in embrace," "copulation," and "a mother holding a child" (Mellaart 1963:148; Todd 1976:93).
2. The item presented in Figure 9.29:b has been described as "bird design at various stages of conventionalization" (E. F. Schmidt 1933, Pl. 88).
3. Many of the items presented in Figures 10.11-10.14 have been described as representations of "the great goddess" (Gimbutas 1982, 1989).
4. The item presented in Figure 10.10:c, and other similar items, have been described as "birth-giving goddess" (Gimbutas 1982:176).
5. The depictions presented in Figures 11.3:e and 11.6:a have been described as "a combat" or "victory scene" (Petrie 1920:16; Williams 1988; Hendrickx 1996; Dreyer et al. 1998).
6. The item presented in Figure 11.26:b has been described as a "town wall with sentries" (Hoffman 1979:148).

By contrast, some scenes from Çatal Höyük that have been previously described as representing dance are rejected here (Fig. 2.19; see discussion below). Decoding meaning from art objects is a basic problem one faces when dealing with items from the past. Scholars from the discipline of art history have written volumes on this subject (see, for example, Panofsky 1955; Gombrich 1972; Bryson 1983). The main working tool at our disposal for this purpose is iconography, which is a branch of art history that concerns itself with the subject depicted in works of art (Panofsky 1955:26). As most of the works by art historians deal with items from historical periods, where religion and mythology are well documented, their methodology is not very helpful when prehistoric material is considered. Nevertheless, the following warning is no doubt relevant to all:

One methodological rule, at any rate, should stand out in this game of unriddling the mysteries of the past. However daring we may be in our conjectures--and who would want to restrain the bold?--no such conjectures should ever be used as a stepping stone for yet another, still bolder hypothesis. We should always ask the iconologist to return to base from every one of his individual flights, and to tell us whether programs of the kind he has enjoyed reconstructing can be documented from primary sources or only from the works of his fellow iconologists. Otherwise we are in danger of building a mythical mode of symbolism, much as the Renaissance built up a fictitious science of hieroglyphics that was based on a fundamental misconception of the nature of the Egyptian script. (Gombrich 1972:21)

Dance experts argue about reliable methods for identifying dance versus other possible bodily activities (Hanna 1979). If there are problems with recognizing dance when it is actually in progress, problems with recognizing depictions of dance are even more complicated. Can one suggest objective criteria for the identification of dancing scenes? This issue should be tackled on both the theoretical and the practical levels:

• The theoretical level. As already noted above, dance is created out of culturally understood symbols within social and religious contexts. Are we capable today, after thousands of years, of recognizing and understanding all the cases in which dance is expressed in art scenes?
• The practical level. The talent and ability of the individual artist to express the motif, the objects' state of preservation, as well as the style and raw material used affect our ability to identify a depiction as representing dance.

To overcome these and other problems, the following thirteen methodological remarks are in order.

1. Dance, dancing scenes, and the decorated items

The first point, which needs to be clarified methodologically, is that three totally different aspects are implied in this study:

• The dance. Dance is a cross-cultural phenomenon, which has been observed in every human organization. In traditional communities it bears various social and religious functions.
• The dancing scenes. This is an artistic motif, which displays the activity of dance. In some periods, dancing scenes were very popular, while in others they were neglected. As can be seen from the San Bushmen of Southern Africa, there is no direct correlation between the importance of dance in daily life and the importance of the dance motif in the symbolic expression of the same communities (see below, Chapter 3).
• The decorated items. An object decorated with a dancing motif had a function of its own, over and above the dance and the dancing scene. Moreover, alongside the items decorated with depictions of dance, identical items appear in the archaeological record, which are decorated with other motifs or left undecorated. Thus the specific lifespan of items decorated with dancing scenes deserves a close look, which Kopytoff (1986:66-67) describes:

In doing the biography of a thing, one would ask questions similar to those one asks about people: What, sociologically, are the biographical possibilities inherent in its "status" and in the period and culture, and how are these possibilities realized? Where does the thing come from and who made it? What has been its career so far, and what do people consider to be an ideal career for such things? What are the recognized 'age' or periods in the thing's "life" and what are the cultural markers for them? How does the thing's use change with its age, and what happens to it when it reaches the end of its usefulness?

Most of the items decorated with dancing scenes are elaborate objects: fine pottery vessels, stamp seals, and cylinder seals. These were commodities manufactured by specialized potters or seal-cutters. Kopytoff (1986:64) has suggested an approach to them:

From a cultural perspective, the production of commodities is also a cultural and cognitive process: commodities must be not only produced materially as things, but also culturally marked as being a certain kind of thing. Out of the total range of things available in a society, only some of them are considered appropriate for marking as commodities. Moreover, the same thing may be treated as a commodity at one time and not at another. And finally, the same thing may, at the same time, be seen as a commodity by one person and as something else by another.

Therefore there should be a clear division between the items decorated with dancing scenes, the dancing scenes, and the dance itself; each subject has to be analyzed separately. The structure and function of dancing activity are analyzed in Chapters 2 and 3. The aspects related to the depiction of the scenes and to the objects on which the motif was depicted are presented in Chapter 4.

2. Depiction of movement in a static medium

When dealing with the depiction of dance in painting, plastic representation, or carving, a fundamental difficulty arising from the subject itself should be taken into account: Dancing is, in and of itself, a dynamic activity performed at a certain time and place, while painting and carving are, by their very nature, static. How can one document movement in a static artistic medium (Bellugue 1963; Arnheim 1974:403-409; Watts 1977:76-86; Gombrich 1982)?

Documenting dance is a difficult task: "An adequate system of dance notation, which is one kind of permanent record, must deal successfully with three different elements: movement through space, movement through time, and the stylistic variations and idiosyncrasies that comprise what we may call 'performance'" (Royce 1977:39). These difficulties exist today, even with the development of modern choreography (Hutchinson 1954, 1984, 1989; Eshkol and Wachmann 1958) and the use of advanced filming techniques (see, for example, Royce 1977:54-55; V. L. Brooks 1984; Morphy 1994). These difficulties were much more severe when the techniques available were painting pottery vessels or engraving stone objects. There was no intention of notating dance at all, and it seems that the aim of a dancing scene was to provide a memory-aid for those who already knew the style and sequence of the movement. Indeed, in many of the items presented below, no attempt was made to represent dancing situations realistically. On the contrary, the artists pursued a minimalist approach, using the three stylistic categories defined below as naturalistic, linear, or geometric. All three approaches concentrate on expressing a small number of characteristics that represent dance:

• The circle. In the archaeological documents there is a clear preference for choosing round objects for portraying dancing scenes. Pottery vessels and cylinder seals constitute ca. 93 percent of the sample presented in the figures (see Table 6.1). Only in a very few cases were other objects used, such as rocks, stone slabs, floors, walls, stamp seals, or a linen shroud (Figs. 7.3:a, 7.6, 7.9:c, 11.24, 11.26:a, 12.2, 12.5, 12.9). On pottery vessels and cylinder seals, the figures were arranged around the perimeter in one line, parallel to the rim, not at random or freely all over the surface. Whether dozens of figures are represented, as on vessels from Tell Halaf or Samarra, or only two are present, as in the examples from 'Ein el-Jarba and Ismailabad, they are organized in a circle around the vessel (Figs. 8.10:a, 8.22:a, 8.30:a, 9.17:e). Arranging the figures in this way creates a symmetrical pattern, usually described as an "endless motif." When the components are geometric, zoomorphic, or floral, the arrangement is merely decorative (von Wickede 1986). However, when the "endless motif" is composed of human figures presented in a dynamic posture or holding hands, it cannot be considered as merely decorative any longer but represents a circle of dancing figures. Moreover, sometimes the dancing figures were drawn on the inner rim of jars, where they are not easily visible. Other figurative motifs were never drawn at this location. These cases emphasize that the dancing figures convey a more complex message than the zoomorphic or floral endless decoration.
• Direction of movement. In the archaeological documents, the movement of the dancers in each scene is always uniform, and all figures always turn in the same direction, thus strictly regulating the direction of movement.
• Rhythm. Rhythm, so essential to dance, is achieved in each scene by the arrangement of the figures at constant distances from each other.
• Body position. In many examples of the naturalistic and geometric styles, the dancers' bodies are represented with the limbs bent, first horizontally and the further limb part vertically, with the arms turned upwards and the legs downwards (see discussion below, Fig. 2.4). This is not a static standing or sitting position but rather a dynamic one requiring great effort. In addition, despite the schematic representation of the figures, fingers were depicted as well in some cases. Hand and finger movements constitute an important component in various forms of dance, and this may be the reason for their inclusion in the dancing scenes, despite that the usual absence of fingers in schematic depictions of human figures.

3. Depicting the richness of dance in restricted media

Dance is usually characterized by a plethora of details: the position of the body and limbs, various articles of dress such as hats or coiffures, ribbons at the back of the neck, belts, and special shoes. Some figures appear to be wearing masks. The drawing or carving of such details, particularly when executed on a miniature scale, are inevitably limited. How can one document the richness of dance with extremely limited means of expression? The artistic solution in this case, as can be seen from the examples analyzed in this work, was to select only one or two elements for a specific scene. Many details have been deliberately omitted, so that only a segment of the entire dance activity and dance decoration is expressed in the depictions.

In sum, despite the difficulties inherent in representing dance, an artistic concept developed in the ancient Near East that succeeded in conveying a complex message through minimal means of expression.

4. Stylistic analysis

In this work I do not undertake a thorough stylistic analysis, as is usual for art historians. However, sometimes the representation of dancing is so schematic that a stylistic analysis is necessary to clarify the depiction (Chapter 8). The painted items from the ancient Near East are classified here into three basic styles: naturalistic, linear, and geometric. There is a gradual transition from one style to another, so that it is not always possible to relate an item to a specific style. The graphic solution to this problem is given in Figure 8.8, where the different styles are represented in a circle.

The basic approach adopted for understanding the schematic styles is to follow the gradual transition from items depicted in a more realistic way to items depicted in a linear or geometric style. To maintain a convincing interpretation, it is important to find both realistic and schematic items at the same site. When both styles are discovered together they create a smooth interpretation, as one can contribute toward the better understanding of the other (Garfinkel 1993:125); the naturalistic depictions help us understand the schematic, linear, and geometric items, while the schematic items indicate which details in the naturalistic depictions are more important than others. These more important details have survived through the process of schematization, while the less important details have disappeared. A hierarchy of details (symbols) can be obtained by comparing naturalistic with linear and geometric items.

5. Graphic presentation

When the depiction of dance was executed on a flat surface, such as a wall, a floor, or a stone slab, it can be easily drawn on a piece of paper. But when the dance was depicted on a round object, such as a pottery jar or a cylinder seal, it is a much more complicated situation. Technically, to draw the three-dimensional rounded item on two-dimensional paper, we need to break the circle and to make a decision about where to start the scene. Such a decision, however, is not only technical but has direct influence on the understanding and interpretation of the scene (Garfinkel 2001b).

Most of our examples were made on pottery vessels and cylinder seals, and on the practical level, the rendering of it on paper could be a real problem. However, usually the dancing scenes are symmetrical, and thus their drawing can be taken from any point on the item's circumference, without making a difference to the result. But when the scene is not symmetrical, different drawings can be made. In this study this problem is relevant to a number of items, all from Predynastic Egypt. Two different presentations have been offered to a vessel from Umm el-Qaab (cf. Fig. 11.6:a with Dreyer et al. 1998, Fig. 13) and three different arrangements to the depiction on the vessel from Brussels (cf. Fig. 11.6:b with Baumgartel 1947, Fig. 14 and with Williams 1988, Fig. 35).

6. The technique and materials

Dancing scenes were depicted on a large variety of materials using various techniques: engravings on stone vessels or stone slabs, applied decoration on pottery vessels, incision on pottery vessels, painting on pottery vessels, painting on walls, and engraving on stamp or cylinder seals. The material on which the scene was depicted affected the final product. Painting on pottery vessels is a delicate technique, which enables clear expressions of the dancing scenes. Usually, naturalistic painted pottery items have been identified by the archaeologist who unearthed them as representing dance. The situation is different with scenes engraved on stone or produced by applied decoration on pottery vessels. Most of the nondancing interpretations mentioned above ("copulation," "a mother holding a child," "the great goddess," "birth-giving goddess," and "town wall with sentries") are related to such objects.

The type of material used has social implications. In the earlier periods, most of the scenes appear on pottery vessels that are the products of rather popular craftsmanship. In later examples from the Near East, large numbers of scenes come from cylinder seals that were manufactured by professional seal cutters.

7. State of preservation

When dealing with items nine thousand to five thousand years old made on breakable materials such as pottery vessels, we do not usually have complete objects at our disposal. In many cases only fragments have been preserved. Since this study concentrates on dancing scenes, a single dancing figure on a pottery sherd raises a problem, as it may imply two different situations. On the one hand, these items could come from vessels bearing several such figures and could thus be fragments of a larger dance scene. On the other hand, these items could come from vessels on which only one figure was depicted and would thus not be fragments of a dance scene. Items of the first type may be included in the assemblage, while items of the second type would not normally be included. However, almost all examples of complete vessels or large sherds that have been preserved show two or more identical dancing figures (see, for example, Figs. 7.10, 8.4:a, 8.10:a, 8.29:b, 9.29:d). Very few complete vessels depict a single dancing figure, and in most of them the other side of the vessel is not shown in the publications, so that it is not clear whether additional figures were depicted on them (Figs. 10.9:b; 10.13:e; 10.18:e; 11.3:b). Because of the ratio between examples where several figures are depicted on a vessel and the few cases where only one dancing figure is depicted, I have decided to include sherds with a single dancing figure in this study. Thus, all relevant data on early dance is included here, creating a complete picture. In addition, the examples of complete pottery vessels with a single dancing figure on them are also presented here, since they are so few.

8. The relevance of the dance displays to the study of dance

Unlike ritual paraphernalia, which is a direct product of rituals and religious ceremonies, dancing activities do not leave direct evidence in the archaeological record. The dancing scenes we have in hand are an artistic transformation of reality into depiction. This situation has been described by Longstreet with regard to other periods, but the principle applies equally to the early items of this study:

The artist is always looking for the one set of lines, the proper curves to bring to life the lightness of the spirit that animates a body of solid bone and muscle as it moves to music. In the dance the body takes on a rigid discipline, a controlled series of lessons that must match those of a partner or even a roomful of dancers. The skill of the eye and fingers of the artist does not merely try to capture reality--for reality is far from the purpose of the dance--but the impression, the fleeting stop-motion of an intricate spin or turn. . . . The dance becomes the memory of a line, a form, a color; and a drawing an impression, an expression, usually abstract (Longstreet 1968).

Thus, are we allowed to see the dancing scenes as presenting some sort of reality, or is the entire depiction nothing but artistic convention, an aesthetic approach (on the aspect of aesthetics, see Taylor et al. 1994), or the result of cognitive processes? In a similar fashion, the student of Greek dance is faced with certain restrictions when evaluating the material. As Lawler has discussed:

Archaeological sources are of prime importance to the student, and serve to render the dance strikingly vivid. But, on the other hand, no sources are so capable of serious misinterpretation. In the first place, they usually have come down in a more or less damaged condition. In the second place, the student must never forget for a moment that Greek art is often deliberately unrealistic, and is concerned with ideal beauty, design, balance, rhythm, linear schemes, and stylization, rather than with an exact portrayal of what the artist saw in life. In the third place, the observer must understand and allow for technical limitations, especially in the work of a primitive artist, and for artistic conventions found in each of the arts, throughout the whole span of Greek civilization. These are not easy facts for the amateur to grasp, and a great many amazing errors have been made by writers on the dance who have tried to interpret representations in Greek art without knowing how to do so. The results are sometimes as absurd as would be similar attempts to interpret modern art realistically (Lawler 1964:17-19).

Taking into account these limitations, it is clear that not every question concerning dance can be answered. Only a rough general outline can be reconstructed, while the small details are not retrievable. To verify the relationship of the dancing scenes to the study of dance, it is important to see who produced the items decorated with depictions of dance and who was the intended audience for these scenes. Most of the items in the early village communities were pottery vessels made by local potters to be used by members of the same communities. For the dancing scenes to communicate a message, they should have delivered a meaningful image, that is, be understandable to the inhabitants of the early villages. This means that the dancing scenes were comprehensible representations of dance activity. They are therefore more authentic than photographs of dance taken by modern western anthropologists in traditional societies. Since the dancing scenes reflect inside knowledge, they may be considered reliable representations of their time. Unlike ritual paraphernalia and although they are not a direct product of rituals, it is justifiable to consider the dancing scenes as authentic documentation of dance activity. Undoubtedly, only a small part of the richness of movement and decoration is recorded in them, but a fragmentary picture is the basic quality of any subject investigated by archaeologists and historians.

9. Categories of dance analysis

Dance activity is composed of individuals who are acting together. Thus there are two levels of analysis: the individual and the community. The analysis should begin on the individual level, which is the basic unit producing dance, and examine two aspects:

• The position of the body. The arms and legs in particular indicate the dance gesture.
• Dance decoration. The paraphernalia accompanying the dancers may include different body ornaments or supplementary objects.

The analysis should continue on the community level. Analysis on the community level takes into account aspects related to the interactions created during dance:

• Dance form. The figures' interaction with the space creates the three basic dance forms: circle dance, line dance, and couples dance. This aspect can be retrieved from the location of the figures on the vessel.
• Dancers' interaction with each other. This aspect examined the figures' interaction with each other within the dance form. The basic compositions are freestanding, holding hands, shoulder to shoulder, and embracing.
• Direction of movement. This aspect can usually be determined when the figures are presented in profile. Two basic directions of circling are possible: clockwise movement and counter-clockwise movement.
• Gender. Gender analysis adds another dimension to the community organization by demonstrating who dances with whom. There are three basic situations: male dance, female dance, and a common dance of males and females together.
• Place of performance. Sometimes some information is given concerning the location of the dance. These depictions include architectural or other elements in whose vicinity the dancing takes place.
• Time of performance. This section examines if there are any indications that the dance was carried out during the daytime or during the night.

The analysis of dancing scenes can be supported by the discipline of dance research as an independent field of study. There are dance academies, departments of dance in many universities, scientific journals for the study of dance, as well as hundreds of monographs on the subject. Beside numerous case studies, there are specific works that have been devoted to a more general level of dance analysis (see, for example, Sachs 1952; Rust 1969; Lange 1976; Royce 1977). This rich reservoir of resources should be integrated in any study that deals with depictions of dance. However, one has to be aware of the specific socioeconomic milieu when analyzing depictions of dance with the help of modern dance research.

10. Assemblage analysis

The focus of analysis should be the assemblage of dancing scenes rather than individual objects. Each of the isolated scenes, when examined alone, cannot produce much information because of the above-mentioned problems of expressing the richness of dance by minimal schematic means of expression. Only the combination of all the relevant data together, when bits and pieces are placed alongside each other, can produce a meaningful analysis. A critical mass is needed to produce a multiple effect.

A major problem here is how to define the assemblage. Our case study considers dancing scenes from western Pakistan in the east to the Danube basin in the west, with a chronological duration of some five thousand years, presented together. The following points justify this presentation: There is a geographical continuity in the distribution of the dancing scenes. There is a chronological continuity in the duration of the dancing motifs. The human societies within these chronological and spatial boundaries were undergoing similar socioeconomic processes of adopting an agricultural way of life and agglomerating large communities into villages.

My work is thus dedicated to the study and analysis of dancing scenes in the context of early village communities. The social context is the major justification for the boundaries of this case study. It would be equally legitimate to investigate the evidence of dance scenes in hunter-gatherer societies or the depiction of dance in urban social organizations.

The motif spreads from a core area, the Levant of the eighth millennium BC, to neighboring regions: northern Mesopotamia and Anatolia. To the east the motif spreads to southern Mesopotamia, Iran, and western Pakistan. To the west the motif spreads to the Balkans, Greece, and the Danube basin of southeast Europe. The motif may have developed simultaneously in different regions. However, since we have chronological and geographical continuity, diffusion is the better explanation.

From a practical point of view, when more dancing scenes are unearthed during fieldwork, it will be possible to subdivide the analysis into smaller regions. Each one of the six major chronological-geographical units discussed here could become an assemblage in its own right. It would be of interest to compare these units and to trace processes of continuity versus change over time and space in them. At this stage of research, however, each unit is statistically too small to maintain an independent assemblage (excluding Predynastic Egypt in a few regards).

11. Quantitative analysis

Statistics should be applied when possible as the chief research tool with which to analyze an assemblage composed of a large number of individual items. The quantitative dimension helps clarify what is more important and what should be considered as marginal in the dancing scenes.

On the practical level, however, there are various problems when applying statistics to our dancing scenes. Many items are broken, and so much data is missing. In addition, some of the variants, such as body gesture, direction of movement, and gender, are not clear-cut categories, and so subjective judgment is involved. Nevertheless, some important trends emerge from the numbers in Tables 2.1-2.4, of which otherwise we would not be aware.

12. Contextual analysis

Another way of understanding dancing as well as the dancing scenes is the contexts with which the depiction of dance are associated. The two most relevant contexts are the following:

• Type of object. This context examines the object on which the dancing scene was depicted. As we shall see below, in most of the cases, there were pottery vessels, small and medium-sized tableware used for serving food.
• Place of discovery. This context can contribute to our understanding of dancing as well as the function of the vessels that were decorated with dancing scenes. Items decorated with dancing scenes were found both in domestic contexts (dwellings units and garbage pits) and in ritual contexts (public buildings and grave goods). Information concerning the place of discovery is not always available. In fact, in most of the preliminary excavation reports and in a large number of final publications, this information is missing. However, in some cases it is possible to see differences between dancing scenes discovered in graves and dancing scenes discovered in settlements elsewhere.

13. Boundaries of interpretation

While dealing with the structure of the dancing performance and with the social function of dance activity in this work, I will not try to interpret the content of the dance. From ethnographic observations, the following occasions for dance have been noted: medicine dances (healing), fertility dances, initiation dances, marriage dances, funeral and scalp dances, war dances, and astral (moon, sun, rain) dances (Sachs 1952:62-77, 124-131). Ackerman suggested that the depiction in Figure 9.5:a represents a lunar dance (Ackerman 1967). Various scenes from Predynastic Egypt have been connected to "great fertility-goddess," "fertility rites," "sacred marriages," and a "cow-dance" (Baumgartel 1960:144-147). Alexander Marshack suggested to me that some scenes, in which the figures hold branches, describe agricultural rites (personal communication). It is probably possible to raise other suggestions as well. In this study some observations have been made concerning depictions unearthed in graves and their relationship to mourning dances (Chapter 2). However, my current impression regarding this type of enquiry, labeled here "content analysis," is that one cannot achieve a reasonable level of certainty. Maybe in the future, with the development of a proper methodology, this type of analysis will bring more fruitful results.