Dagobert Frey (1947) and Otto F. Bollnow (1963) have both elaborated on non-homogeneous or polar space concepts - the former on sacred architecture throughout the Afro-Eurasian belt of ancient high cultures, and the latter phenomenologically and anthropologically. Besides these two innovative works there seems to be very little study on the evolution of human conceptualization of space, in spite of the strategic meaning of space conceptualization in various disciplines. If, with Bollnow, we assume that space conceptualization was originally related to settlement, then most of what is translated of ancient texts or what has been reported in ethnography would have to be reconsidered.
According to Bollnow, human space conceptualization evolved through time from small dimensions to large extensions. This means that there were originally only a limited number of types which could be studied objectively within the human environment. In our opinion, the polar or complementary type is the most important - and the most widespread. Many social, aesthetic, religious, and temporal expressions of quite different cultures can be explained by a 'threshold' model which implies a semantic element situated in the linear center of two categorically opposed spatial domains. This model corresponds exactly to the structure and flexibility of this type of space perception. If, consequently, we assume that cultural or anthropological space conception was topologically or environmentally rooted and extended in polar relations  around the topos, this would imply the following consequences: sites, sub-sites, faces, and sub-faces would have to be recorded-as a whole with exact localization of all figures within the whole of the arrangement. Each figure, even if not identifiable, assumes a potential meaning within the whole. A given set of sub-figures, figures, or a site or a sub-site would have to be checked against the hypothesis that it mapped some close relation with the spatial organization of a temporary or permanent settlement, its social structure and its territorial conditions at a certain time. Hypothetical typologies of settlement-space organization as coherent system applied to a site would imply a need to consider its relation to other topological and environmental conditions, including the wider tectonic conditions such as mountains, rivers, stretches of water, etc. Testing the hypothesis that topological arrangements used the access-place-scheme, sites or sub-sites (e.g., cave) would have to be thoroughly analyzed with respect to the corresponding criteria (Fig. 34).
In general, the spatial should become a more important criterion in petroglyphic interpretation. It might become critical to ascertain where, in a given setting, a particular drawing is found. Sites of petroglyphs would have to be analyzed as a spatial system where the rear, center, and the access region are characterized by certain categorial values (static/dynamic; defined/ill-defined, artificial/natural, etc.) which may reflect on the formal conditions of signs, drawings, or tectonic design. Following Bollnow's theory of the 'micro-origin' of space perception, if the cosmic dimensions were relatively late factors in cultural evolution, we would have to ask what predated them. From this point of view the following hypothesis may be useful for rock art research.
The river was an important system of orientation in early settlement. The local river is not only of survival importance in regard to water supply, or in other essential economic ways (e.g., fishing), but implies a directional element from above (mountains) to below (ponds, lakes, sea) and, with branching rivers, provides an environmental network which allows localization and evaluation according to sources and mouths, according to upper and lower domains of a river.  Figure 35 shows a reconstruction of the 'horizontal cosmos' of the Ainu. The principle of 'coincidence of opposites' sets their sacred signs (S), their dwellings (D), and their territory (T) into analogy or homology. In the philosophical world-view of the Ainu, these different components are identical. All are dealt with in the same way, with the same intention to harmonize their environment. [l3]
For archaeology and prehistory, the excavation and interpretation of settlements is crucial. But the reconstruction of the vital structure or the meanings of such settlement-remains is often extremely difficult. All that could explain former meanings is lost. History and folklore, on the other hand, have a relatively good knowledge of rural settlement patterns, and know rural life throughout the European continent quite well, but history in this European domain was extremely dynamic. The historical or vital typologies are difficult to apply to prehistoric conditions.
Unfortunately, ethnology has very much neglected this important field of settlement studies. The emphasis has been on mobile material culture, on collections of easily transportable objects for the museums at home, and on social anthropology, religion, economy, etc. Only gradually is it being realized, mainly by forces coming from recent architectural research,  that the settlement studied ethnographically in its objective, spatial, and temporal aspects depicts a vital complex which questions, in many ways, conventional approaches. It becomes evident that the Western analytical approach developed in European urban conditions proceeds in facetted disciplinary divisions, and thus distorts or dissects, rather than factually represents, what counts: the vital and quite reasonably functional complex of the settlements past and present, as represented in the traditional rites and cults. 
If approached ethno-(pre-)historically with a wide horizon, comparative settlement research could mean considerable progress for archaeologists and prehistorians. Typologies of settlement patterns and space concepts would be valuable at various dimensional levels: if types are limited and predictable within tectonic landscapes, this could contribute to finding new locations of prehistoric remains. A wider opening of the comparative research dialogue of ethno-prehistory would not only yield more realistic clarifications; we might also discover that our concepts of 'primitive' hunter and gatherer bands are quite illegitimate reductions, projections of our very limited point of view. Our study of the Ainus -originally a society of food-gatherers and hunters in the North of Japan -showed that this so-called primitive ethnic group had an admirably developed system of environmental organization based on the concept of the 'coincidence of opposites'. This philosophy structured the manner of their work, the evaluation of their landscape, their festivals, their houses, and their way of living in them. And this whole 'metaphysical' world-view was formally embodied in their age-old tradition of (perishable) sacred signs.
Recently, the interest in 'stony signs' has increased considerably worldwide. This is certainly an interesting development, and not only because its sources are quite different from the usual archaeological excavations. Scratched or painted by humans on durable surfaces, these designs correspond more to a prehistorical museum and thus transcend the notion of mere 'remains'. We can assume an active personal or impersonal will to represent on the surface used a part of the factual environmental reality. Rock art gives us a human 'view', much more so than what has been preserved in the ground. Consequently, deciphering this material involves much more and certainly quite different methods than merely interpreting found 'leftovers'.