Based primarily on the ideological biases of historically founded 'high' religions. the Eurocentric and hierarchically structured systems of religions created a category of 'lower' religions or 'primitive' beliefs, particularly if such cults were related to primitively constructed objects. Though 'fetishes' and 'idols' played an important role in earlier theories of religion (de Brosses 1760; Comte 1851; Tylor 1871; Schultze 1871; Frazer 1890; Van der Leeuw 1933, 1948; Mannhardt 1963), the outlook was dominantly animist, emphasizing the 'belief-systems'. Consequently, such objective representations (semantic architecture) have not been duly studied in their material, constructive, formal, or spatial aspects.
Paradoxically, the history of Christianisation has documented globally quite well the existence of such 'sacred' objects, because the theologically educated missionaries used the artefacts of traditional cults as evidence to stigmatize the 'primitive' character of such beliefs, thus justifying their own efforts to 'civilize' and convert the populations which produced them. The geographically and historically universal distribution of 'semantic architecture' is very important because it allows us to discuss the phenomenon anthropologically and to construct a framework of possibly global validity. Naturally, such sacred objects could not be found as archaeological remains because of their ephemeral material quality. On the other hand, numerous archaeological iconic objects made of durable materials suggest the prehistoric existence of semantic architecture, and various other indications lead us to the assumption that semantic architecture was widespread, particularly in neolithic times when it was an important and central element of village cultures (Fig. 7).  (For details see Egenter 1984.)
Research into semantic architecture is part of a wider framework, namely architectural anthropology, which we will describe briefly to show that the following suggestions are not isolated but stand in a global systematic framework.
Architectural anthropology is based on a new, comprehensive definition of architecture. It includes everything constructed by Homo sapiens sapiensand his predecessors. It is a theory of architecture dealing with architecture and buildings in the wider sense of constructive behavior and its effects on the human perception of environment and space (Egenter 1992). There are four classes based on factual source materials: (1) Subhuman architecture: nestbuilding behavior of higher apes. (2) Semantic architecture: non-domestic buildings or buildings unrelated to the human body endowed with semantic, social, and ideological functions. (3) Domestic architecture: buildings which offer space for protection of objects, animals, and humans. (4) Sedentary architecture: higher, horizontally structured units assembling several elements of semantic and/or domestic architecture.
Thus architectural anthropology is based on a particular. approach, focused inductively on objective sources. In its reconstructions it does not rely on the historical method, which it considers too speculative. Against the classical archaeological method it postulates that what was most relevant for cultural evolution was not made of durable material. Against history in the narrower sense, insofar as it deals with the early history of any culture, it maintains critically that the most important dimension of societies in the proto- and early historical phases - that is to say, what is generally called religion - is likely to be misinterpreted by means of historical (scholastic) retroprojections (Egenter 1992). It also contends that most interpretations of early texts show a very uncritical use of space concepts and thus legitimate an 'irrational' condition of thought. Important studies, such as that of Kerschensteiner (1962)  on the meaning of the term 'cosmos' in ancient Greece, are not sufficiently taken into consideration. What the historians call 'world-creation-myth' can often be traced back to the class of very locally limited settlement foundations (e.g., the Babylonian creation myth as foundation of a settlement, in Winckler 1906). 
Methodologically speaking, architectural anthropology is based on what the Vienna School calls 'structural history' (Wernhardt 1981 ), which questions the epistemological value of history based on the sequence of dated findings. In contrast to the historical method, it emphasizes interdisciplinary methods like ethno-history, and ethno-archaeology, and ethno-prehistory, thus searching for meaningful structural contexts. Since architectural anthropology searches for patterns of constructive behavior rather than its remains, its focus is on the ethnographic evidence provided by traditional cultures which are still observable. Its goal is the systematic reconstruction of a constructive continuum, one which parallels the whole process of cultural evolution, including early phases of hominization, prehistory, history, and the present.
The understanding of subhuman architecture (nestbuilding behavior of the higher apes) is essential to this approach. As early as 1929, Yerkes postulated nestbuilding as the evolutionary basis of human 'constructivity'. In our concept it forms the firm scientific ground upon which it is possible to define and reconstruct architecture anthropologically (see Egenter 1983, 1990).
Japanologists consider Shinto as the autochthonous religion of the Japanese archipelago. Historically, it was superseded by Buddhism at the beginning of the eighth century in central Japan. In urban centres Shinto was assimilated in part by Buddhism's historically founded religion and temple systems, but has preserved a purely traditional form in its wide-spread village cults of the protecting local deity (ujigami). These often isolated village traditions clearly show traits which are deeply rooted in the prehistory of the Japanese archipelago. Japanese pre- and postwar folklore research (Yanagita Kunio) has collected a valuable corpus .of data, particularly on rural Shinto cults and rites. But these important ethnographical sources have largely been neglected by Western researchers - first, because Japan is considered to be a high culture, and thus folklore studies are thought to be of secondary importance; and second, Japanese folklorists have followed Western schools in the interpretation of their own materials, and their particular significance has been over-looked. Naturally, another likely reason for this neglect is the language and script barrier.
Many traditional rites of Japanese village Shinto include or even have as their central content the construction of semantic architecture: pillars, huts, and other architectonic forms. But a wide range of bio- and techno-morphous forms is also observed. Such signs and symbols are constructed with primitive methods and are set up in particular places within the spatial structure of the villages. In order to define our inductive architectural approach, we must suggest some limitations to the conventional classification of such Shinto rites.
In religious terms, these signs and symbols are considered to be temporary 'seats of gods' (yorishiro). According to official Shinto religion, the spirit of a historically deduced deity descends from heaven into the sign and presides over the sacrificial ceremonies and banquets of the celebrating community. After the destruction of the sign, the sacred spirit ascends again to heaven. However, this theological interpretation does not reflect the actual original practices. Field research shows clearly that this exegesis of the ritual must be considered a secondary interpretation which spread to the villages with the official Shinto theology developed from Buddhism. What really counts is the traditional construction of the signs themselves: the rites and the built signs are the message (Egenter 1980, 1982; Ludwig 1983). Thus, the study stresses the conviction that the metaphysical aspects of semantic architectural forms are an intrinsic part of the tradition itself. The cultic value of the signs is based on the technological conditions of the structures and their territorial functions. The technological aspects are indeed important. Semantic architecture in Japanese village Shinto shows three essential characteristics: (1) the constructions are essentially fibrous (vegetal flexible materials are used); (2) the technique implies the hand as a tool; and (3) all materials and processes are ecologically autonomous. These characteristics can be considered as criteria of autochthonous origins, indicating that this tradition may be deeply rooted in prehistoric times, since the materials and the constructive techniques used are consistent with very early forms of culture (Figs. 8-10).
Thus we can recognize a deeply rooted global parallelism between such types of signs and symbols. It is evident that, proceeding from a relatively differentiated human constructive capacity, this type of primitive industry could have emerged anytime and anywhere, given certain environmental conditions. But there is another important aspect related to this objective semantic system.
Philosophically, these structures are very difficult to describe. Materially and in their outlines they form a unity but, if -their qualities or categorial characteristics are taken into consideration, they are dual or bipolar. If the rope is included as an element, we would have to speak of a 'trinity' in which the rope is the means to achieve the whole form, its condition .sine qua norm. The form is essentially non-logical or 'irrational' (Fig. 16). But if we look at the scheme which represents different forms all showing the same basic structural principle, we find a model for a cognitive system which may include quite different forms, identifying them as a higher group by their analogous or harmonious structure (Fig. 17).
Thus the local distinction (each settlement has its own form) is inter-woven with a structural system (all forms are harmoniously structured) which allows identification in spite of differences. Two cognitive aspects of the same forms are combined: a differentiating and a generalizing one. In regard to early conceptualization, it may be remarkable that the pragmatic territorial function is related to the individual characteristics, whereas the structural symbolism, the spiritual and irrational element is related to generalization.
Meaning follows categories of ritual use. There are many cases where such mobile pillars have to be moved from one significant place to another. In the ritual context they obtain a markedly dynamic character while on their way. As a category, this dynamic phase is markedly opposed to the former topo-tectonic stability. Obviously, this led to metaphoric accumulations of later meanings: the pillars become wild fire-dragons, drunken lantern-ships, or strangely formed fish (Fig. 18). Indeed, the same object can have entirely different meanings according to the way it is ritually used. This means that the object is defined only superficially by its outer form, but essentially by its ritual function which is expressed in categorial terms (places and paths, stability and movement).
Particularly interesting for the origin of symbolic thought is the allusion in this context to the fish and the boat. It is an accretion through analogy: the fish and the boat are the perfect example of horizontal movement in a medium absolutely opposed to stability. Natural and instrumental, the fish or boat forms of our tradition are not really fish or boats, they are in fact mobile pillars. Evidently, they are indebted to the geometry of this tradition (Fig. 19), whose development runs from abstract and geometric forms to natural or instrumental ones.  Thus the word 'stylized' would have to be abandoned: the technological factor would be responsible for what we call 'style'.
In the region surveyed, several villages, showing the same basic tradition of semantic architecture, build artificial trees. This provides us with another model to show how this 'structural symbolism' of a human constructive tradition integrates natural phenomena (Fig. 20). The upper part of the tree is formed by inverted brushes stuck into a bamboo structure anchored in the ground. The brushes are formed around bamboo stalks using evergreen twigs of Camelia, and finally the upper part is cut in some places to create an even-looking tree-top (see Figs. 1-3).
Thus the form of a tree is entirely artefactual; the structural traits underlying the appearance remain closely related to the geometrical nature of the other types of semantic construction. The ritual context is exactly the same as in the case of other purely geometric forms; the tree is burned at the end of the rites. The meaning of the tree is of secondary importance with respect to the form of the object, the main formal motive being a semantic differentiation from the related forms of other villages of the region. We have used this example to show that the uncritical identification of petroglyphic forms with natural objects can be highly problematical. There is another example in the same region which shows traits of anthropomorphic representation: two markers, in the form of pillars whose basic structures are identical, are differentiated: one is male, the other female (Fig. 21). Obviously, neither 'gender' nor 'sexual symbolism' is concerned here. The two villages which are of the same origin, have basically the same pillar-type of markers, but had to differentiate the signs with a supplementary image of opposition or harmony: male and female. The differentiation shapes the female 'body' in slightly curved lines, and breaks down the upper part of the reeds by analogy to the common Japanese hairstyle. There are other details which denote gender. However, the essential meaning is not the naturalistic aspect of each form, but their relation. A few formally quite incongruent traits (body shape, hairstyle, knots) are sufficient to suggest the difference and characterize the relation. If we had found these two objects drawn or carved on a rock surface, the possibility of discovering their meaning would have been very remote.
Snakes are also very interesting in this context. In some villages of the region they play an important role. But this has nothing to do with what one normally calls snake-cult (Fig. 22). Probably due to its characteristic patterns of movement, the sacred rope, which in this region is basically a constructive element holding the markers together and, as such, is of prime importance, was considered to be analogous to natural snakes. In two villages, we found interesting cases. For example, illustration (Fig. 22a) shows a female (left) and a male (right) marker of the fixed type. Both have a thick 'dragon' (ryû) coiled around their 'body', obviously related to the main rope of other forms. Most interesting here is the formation of the dragon-heads. By no means do they show any similarity with dragon heads as frequently depicted in Japanese history of art. They consist of a small bamboo-twig, its pointed leaves moving nervously with the slightest movement of the air (Fig. 22a). Figure 22b shows another interesting case: the sacred rope which marks the entrance to the shrine precinct is considered to be a snake, and, since each rope has two ends - the tassels being interpreted as heads - it has become a two-headed snake. When the signs are built, this rope is cut into two pieces, now forming two one-headed snakes, which are wound around the newly made signs, thus 'guarding' them until they are burned (Fig. 22b). Japan's agricultural ritual scene is full of similar examples of snakes derived from sacred ropes. To interpret them in view of the Western concept of 'snake cult' would be to miss their actual content: the 'snake' is an essential constructive part of an ancient local territorial demarcation system, from which it derives its values.
It is usually taken for granted that early humans perceived the natural objects of their environment exactly as we perceive them, but this may be an illusion. How did this increasing capacity for perceiving structured objects come about? As we have attempted to show, the concept of 'semantic architecture' may provide a structural answer.