ARCHITECTURAL DESIGN AT LATE CYCLADIC AKROTIRI
Understanding architectural design is essential in archaeology, because when 'reading' the architectural relics of our past what we attempt to do - consciously or unconsciously - is to reverse the process of architectural creation: beginning from the finished product - and that not in its entirety - to arrive at those initial factors that determined function, form, and structure. Factors, such as technical standards, social structure, religious beliefs, economic status, prestige etc., from which we shall hopefully piece together an overall picture of life in earlier periods of history. But if we are to reverse the process of architectural design we must first try to understand its logic and find a way to approach it in terms of theory. Such a theory is the 'models and variations' scheme constituting the basis of anonymous architectural design in vernacular architecture.
The rich architectural data available at Akrotiri offer plenty of evidence in corroboration of this theory. The aspects of architectural design examined here refer to issues of form and structure. Function, on the other hand, is a matter of inter-disciplinary approach, since architecture alone can not give definite information on this subject.
READING THE ARCHITECTURAL RELICS OF OUR PAST
Up until the end of the last century our only knowledge of the prehistoric civilizations of the Aegean was that bequeathed to us by the writings of the Classical era. Through myths, legends and tradition emerged a picture of powerful states with civilizations highly esteemed yet already remote and alien to the people of Classical times. It is this inherited memory, alongside a Western culture that can easily recognize its ancestors only down to the Roman and Classical era, that constituted the cultural background of the pioneer excavators on Crete at the beginning of our century. And the first discoveries - the remains of the palace Knossos itself - were such as to 'confirm memory': the great extent of the ruins, the chaotic accumulation of space cells of various sizes and forms, their obscure and intricate intercommunication, the numerous and complicated passageways and staircases - all these comprised a picture of confusion, enhanced even further by its association with the myth of the labyrinth.
It is not surprising, therefore, that since Evan's first discoveries at Knossos so many scholars have hastened to describe Minoan architecture using terms such as "incomprehensible, confusing, illogical, irrational, primitive, insane jigsaw, disorderly agglomeration, etc." (see Graham 1969, 234). Such characterizations, however, are entirely absurd, not only for the architecture of a civilization such as the Minoan but for any work of architecture at any time in history. Because architecture is by definition a pure act of logic. It is true, however, that these extreme ideas, expressed so vociferously in the past, have now been surpassed and scholars of prehistoric civilizations give a proper notice to the value of architectural design (see, for example, Graham 1969 and Preziosi 1983).
Understanding architectural design is essential in terms of archaeology, because when 'reading' the architectural relics of our past what we attempt to do - consciously or unconsciously - is to reverse the process of architectural creation, so as, beginning from the finished product - and that not in its entirety - to arrive at those initial factors that determined function, form, and structure. Factors, such as technical standards, social structure, religious beliefs, economic status, prestige, etc., from which we may hopefully piece together an overall picture of life in the earlier periods of history.
But if we are to reverse the process of architectural design we must first try to understand its logic and find a way to approach it in terms of theory; a general and universal theory, as widely applicable as possible. Such theories can be found in the writings of theoreticians and critics of architecture (Kotsiopoulos 1985; Zevi 1972), because the way archaeologists, architects and other scholars 'read' the architectural remains of our past has much in common with the way critics deal with works of architecture, either past or present: they too work backwards, analyzing the result in order to find the initial intention; they too seek desperately for the criteria that will help them overcome the inevitable subjectivity of their judgment; and they too seek to establish a language of communication that will bridge the distance separating the creator from the observer. Our quest therefore, is for a theory that will help us follow the architect at work.
MODELS AND VARIATIONS
How does it all start? In the beginning there is the Idea of Building. Whether domestic or public, the building is conceived in advance in a vague way. This is possible by what Viollet-le-Duc calls 'pathetic imagination' - i.e. recall of other similar buildings constructed for the same purpose in the past. This is a collective memory, to a certain degree, shared by both specialized builders and common people, this facilitating communication. From pathetic imagination it is the architect alone who must then pass to 'energetic imagination', which is the application of logic to the material of pathetic imagination (Valery 1921, Greek translation 1988, 147, quoting Viollet-le-Duc 1858-1872). And logic is the process of thinking, combining, selecting, defining priorities, calculating, and solving problems. Collective memory gives way now to individual judgment, free associations and subjectivity.
And how informative can pathetic imagination be? Though as a whole the Idea of the Building is a vague and indeterminate image, it nevertheless comprises many well-defined isolated details on issues of function, form, and structure. These details are based on selections made by others, in previous generations, that can ensure, through prior experience, the successful adjustment of the building to the conditions of its natural surroundings and the various needs of everyday life. They constitute a set of 'models' or prototypes, transmitted from one generation to the next in the form of tradition. A tradition that can attain the power of law enacted through collective consent and kept through collective control (Rapoport 1969,22).
As the architect proceeds to analyze and specify the initial Idea of the Building, he finds himself confronted with a series of limitations. The problem that arise must be solved through energetic imagination and the well-known models must be adjusted to each specific case through a new process of selections and decisions that will lead to the most appropriate variation of the model. Rapoport (1969) believes that this scheme of 'models and variations' - another way of describing Viollet-le-Duc's 'pathetic imagination'- constitutes the basis of architectural design in anonymous secular architecture. Benevolo, too, writes more or less the same, referring to the set of known parameters that facilitate achitectural design, since there is no need to start the same process of thinking from the very beginning each time Benevolo 1960, 33).
This widely accepted scheme of 'models and variations', therefore, seems to be appropriate for approaching the issue of architectural design at any time in history. Because generally speaking, what happened in the Aegean in prehistoric times, for instance, is not very different from what usually happens in life when engaging in the act of building; especially when we are referring to the pre-industrial era. There has been no basic change in human logic just as there has been no change in the human body and its basic needs, nor in the effect of the environmental factors on buildings and the physical properties of the basic building materials, e.g. stone, wood and mud. It is other factors that are variable, such as social structure and religious beliefs, and it is these that differentiate the architecture of each time and place (Rapoport 1969).
Let us now turn to our specific case study, the Late Cycladic settlement of Akrotiri, where the possibilities of following the architect at work are more than anyone could hope for, for an era as remote as the 16th century B.C. According to the theories presented above, we may well imagine the prehistoric Theran architect - bearing the Idea of the Building in mind and walking around the actual plot of land where it will stand - finding himself confronted with the first limitations: location and climate, size and form of available land, nature of soil or rock, inclination, orientation, are all factors that should be taken into account in connection with the type of building to be constructed. The immediate surroundings impose the next limitations: building within the settlement this would mean taking into account the existing street network and the possibilities of access, ventilation, lighting and drainage facilities in connection with the surrounding buildings. As has been discussed elsewhere (Palyvou 1986), these are factors of major importance at Akrotiri that determine architectural design to a considerable degree.
In the process of adjusting his model to the above limitations, the architect is led to a preliminary general idea as to where the specific building should stand, what is the maximum area it can occupy and what its outer limits can be. Yet the process of architectural design has proceeded even further, because the spectrum, of selections has already narrowed greatly. Where the access to the building will be, for instance, is probably obvious by now since one of the most basic characteristics of the architectural idiom at Akrotiri is that the entrance opens onto the main street network, and preferably at a point where there is a widening of the public open space. Such a decision draws along a whole series of other decisions, since the entrance is related to all other parts of the building in a predetermined manner, according to the type of building in question (information drawn from pathetic memory - i.e. the 'model').
To test all this in a more analytical manner we would need a significant sample of well-defined separate buildings. But since excavation is still in progress at Akrotiri there are, for the moment, many gaps in our knowledge of the exact limits and room-arrangement of different houses. There are, nevertheless, already sufficient indications of the existence of a standard scheme of space-arrangement, involving certain types of rooms interrelated in a specific way. That is, one may detect a typical 'house model', both in form and structure, which appears in several variations within the area excavated to date, while there is at least one building (Xeste 3) so different from this model that it does nothing less than confirm the rule. For the time being we shall not attempt to give a definitive overall description of the 'house model'. We may, however, point out some of its most characteristic features, mostly as examples that will demonstrate the peculiarities of architectural design at Akrotiri.
The entrance in Theran architecture, combined with the main staircase leading to the upper floor, is the most typical example of architectural design based on the scheme of 'models and variations'. The model , in this case, is more than obvious, because the deviations from it are slight and insignificant. This, actually, is not one model but a combination of many different models referring to each separate part - i.e. door, window, walls, floor, staircase, and so on. For each one of these there are such clear specifications as to position, form, materials, elaboration, dimensions, colours, etc., that if a new house was to be built at Akrotiri there would have been little thought as to where its entrance should stand, how it should look, and how it should be constructed.
Another typical feature of the 'house model' is the large room on the upper floor that one enters from the main staircase: it is usually, but not always, an almost square room with a column in the middle. In Sector Delta - Room Delta 1, however, we find in its place a double room connected by a polythyron. The intention is, I believe, the same: namely, to ensure a spacious room. If this is so, then here the variant is so extreme that it almost counts as an exception. Such cases, however, are very interesting because they give a vivid picture of the architect at work; in tracing the reasons for their existence one might discover the criteria that led to the specific decisions and the priorities which prevailed. In this case, room Delta 1 may have resulted from a later conversion of the upper floor: the existing two rooms at ground floor level, rooms Delta 1 and Delta 1a, could not be united into a large square room with a central column on the upper floor, because the overall size would have been too great. The problem was solved by a well-known technical device - the polythyron - best suited to the situation. This means that by analyzing the 'model' into its components we arrive at the assumption that the priority lies in the large space and not the square shape or the central column. Moreover, it is seen that structural elements - the polythyron in this case - commonly connected with a specific type of room may be borrowed for purely technical reasons, denuded of any other significance. The validity of this interpretation can be demonstrated by other means, such as the finds in room Delta 1 compared with those normally found in the rooms with central column next to the main staircase, as well as other constructional details of this room, such as the treatment of the floor and walls.
Once the model is well understood, then it can be used to test other similar cases. if we know, for instance, the type of room to be expected next to the entrance, then we are better prepared in the process of excavation as well as in the final interpretation of the data. It is more probable, for example, that the entrance Gamma 5, 6 of Sector Gamma leads to a room (Gamma 7) of the type discussed above. Though the area is unexcavated as yet, this is apparent from the overall dimensions of room Gamma 7, which fit well the type of large room with central column; thus, a column base is to be expected under the debris. A similar example is the room next to the entrance of the South House; in this case, it is the type and arrangement of windows on the northern facade of the building, as well as the dimensions of the upper-floor room that reveal the interior arrangement.
The model can be used the other way round too: in Sector Beta, for instance, no entrance has been found; yet, the existence of a large square room with a central column on the upper floor (Beta 2) may well indicate that the missing entrance is somewhere nearby. If we combine this indication with information deriving from the 'entrance model', such as the fact that the entrance is positioned at one of the corners of the building, then the missing entrance of Sector Beta may easily be placed in the destroyed area south-east of room Beta 2.
Building technology at Akrotiri is characterized by high-level techniques and standardization of all constructional details (Palybvou 1988). This shows that able architects and skillful artisans, specialized in the art of building, worked on the basis of a well-known system of specifications (models) that defined all the necessary structural details, materials, sizes, elaboration, methods of construction, even colours, for each part of the building: walls, floors, staircases, doors, windows, etc. Manifest in the building technology, therefore, is the existence of an architectural design based on the theory of 'models and variations'. There are, of course, extreme cases of variations or even transgressions of the rule. These are due to many reasons, such as technical restrictions, experimentation or innovation. They may also be due to deficient knowledge of the model or unskilled workmanship, in which case the violation of a rule may sometimes lead to failure and disaster.
To give an example of such an 'exception' let us examine the oblong horizontal window at ground-floor level on the southern facade of the West House. The shape of this window strikes us as odd, not only on the facade of this building but among all windows at Akrotiri. It is true that the 85 windows known to us from Akrotiri to date are easily grouped into four categories, and that this specific window falls into none of them (Palyvou 1988, 310-316). Yet, on a closer examination, we ascertain that the intention was to have a Type B window here, and that practical reasons (level of Triangle Square and height of ceiling) led to such an extreme variation of the model. Moreover, we see that the first priority in the design of this window was its total surface (it fits Type B windows) which is more than natural since the major contribution of a window is the amount of light it admits by its size.
ARCHITECTURAL DESIGN STEP BY STEP
Architectural design at Akrotiri has a peculiar common in vernacular architecture and perhaps characteristic of an early stage of development: it is applied toe ach part of the building separately without any obvious intention - or possibility - of seeing the edifice as a single entity. This, of course, is due to many reasons, such as the very nature of the anonymous secular building, which functions like a living organism with a series of expansions and conversions; the scale of the work, which makes it difficult to supervise it as a whole; and the agglomerative manner of the construction, which admits very few alterations, readjustments or rejections. Moreover, it is interesting to note that a similar idea of design is attested in the art of this period; each part is rendered properly yet not necessarily in accordance with the other parts. "His love of detail often prevented him from visualizing the detail incorporated into the whole ." (Iliakis 1978, 627).
The entrance may serve again as a good example of this. The 'model' of the entrance is clear and simple: it is an oblong room coinciding with a staircase consisting of two flights and two landings, one of which bears the entrance door. A room of such description has an innate symmetrical character. The actual entrance, however, are all non-symmetrical variations of the model. One basic reason for this is the step-by-step application of architectural design. The two narrow sides, for instance, ought to have been of equal size; yet, the one bearing the entrance is usually larger, because it has to accommodate a door of specific dimensions (those defined by the relevant model) and a Type B window (this entrance-window, so typical of Akrotiri, derives from the strict application of another 'building regulation' that demands abundant lighting of the staircase - very much like modern building regulations). The total dimension thus resulting is larger than that defined by the specifications of the staircase itself - i.e. twice the width of a flight of steps plus the width of the middle wall. The latter dimension is retained, however, for the other narrow side of the staircase since it is not within the logic of Theran architectural design to interfere with the dimensions just for the sake of uniformity, symmetry and overall design. A similar case is the difference of width between the two flights of steps, which again has some logical raison d'etre that applies to only one of the two flights, as well as the uneven distribution of steps, that results in a slight distortion of once of the long sides of the staircase.
Another example is the uniform application of a basic and simple building specification; namely, that all windows be placed in the middle of the wall. By applying this rule to each specific case separately we get awkward results, such as the two small windows at the ground-floor level and the large window above (the two rooms on the ground floor have been united into one room at the first floor) which is a disadvantageous combination of openings from the point of view of statics (the loads transferred to the ground by the jambs of the large window meet the openings of the two small windows underneath). One more example is the variety of widths observed in the exterior walls of a building which derives from the fact that the wall is designed to have the width necessary for each one of the rooms separately.
SPACE ARTICULATION AND FLEXIBILITY OF CONSTRUCTION
In a multi-storeyed, densely organized architecture, as is the case at Akrotiri, architectural design is not only obviously necessary but constitutes a difficult and very demanding task. It implies the distribution of functions on two or three levels, and anticipation of an intricate system of horizontal and vertical communication. This is demonstrated very clearly by the great degree of variation that is attested in the organization of the interior of the building, since the room arrangement of the upper floor is quite different from that of the ground-floor level. This differentiation is achieved through an advanced and sophisticated building technology and is based on the interchange between solid parts and void; that is, compact partitions of one level are perforated or eliminated on the other. A basic structural element involved in this is the pier-and-opening partition. as a substitute for a compact wall. This can take the form of a pier-and-door partition (polythyron) pier-and-window partition (polyparathyron) or pier-and-closet partition (polyhermarion). Another element is the isolated support in the form of a column or a pier. One more is the mud-brick partition wall, which plays an important role in the upper levels since it can be incorporated at any point; being of a light construction its loads can be transferred by the floor beams and being quite thin it takes very little space arrangement some of which are based on highly sophisticated and intricate constructions. One such case is the staircase of Xest 3 where the north rubble wall of the upper floor is carried at the ground floor level by horizontal beams that transfer the loads to an adjacent series of vertical wooden surfaces.
All these examples show that the builders at Akrotiri exhibit great skills in their art, both in designing and conceiving what is to be a large multi-storeyed and multi-functional building, and in executing this concept. Long before starting to build, they must have anticipated what the loads of the upper structures would be and how these would be transferred to the ground; issues of great importance and decisiveness from the very first stage of the building operation, that of the foundation,
FROM CONCEPTION TO EXECUTION
Architectural design implies an intricate process of thinking that demands a flexible mind, strong imagination, foresight and heightened abilities in visualizing space in its three dimensions. To make this concept material, however, a special tool is needed: a technique that will allow the architect to transfer the image from his mind to the ground and help him keep control on what he is doing. In the modern sense of an architect at work, one would differentiate between planning (thinking), designing (calculating and drawing) and laying out (executing). But for any period of history before industrialization, vernacular architecture - 'architecture without architects' (Rudofsky 1965) - involves, most probably, all three in one: the building that has been planned is probably designed to a great extent at the same time as it is laid out on the ground; there is, that is to say, a direct feed-back process between all factors at play, giving immediate results and substituting the drawing boards of the modern architect.
The tools he may use for laying out the building may be ropes, pegs, leveling devices, and even some measuring rod; but these are more useful to check if the 'models' he has in mind fit the allotted area and to decide the sort of 'variation' he must apply (the models, we recall, contain dimensional specifications as well), rather than to execute an overall, predetermined, finished design. This process is evident from the fact that there are almost no right angles and no pure geometry in the buildings of Akrotiri. In an architecture of well-defined geometry, on the other hand, a specific design and layout technique are surely in use, because such shapes can not come about incidentally (Preziosi 1983).
Laying out a building at Akrotiri, however, is facilitated by the fact that its basic structure is simple. In most buildings, we observe that the main fabric consists of a plain juxtaposition of space cells one next to the other. It is by these intelligent technical devices - innovation of Bronze Age Aegean architecture - woven into this primitive fabric of the building, that the edifice as a whole becomes a sophisticated articulation of space. One such device is the pier-and-opening partition; and it is interesting to observe how the innate order of this structure inspires the architect to apply it in a wider sense, as in the example of Room 5 on the upper floor of the West House - an excellent sample of the 'ever modern' in architecture. The repetition of a unit - the door - leads to the grid. Yet, this is a latent grid, because there is no strict module as far as dimensions are concerned. Moreover, we see once more the limitations of a design that does not exceed the boundaries of a room. The same applies to the eastern part of Xeste 3, where the numerous polythyra in this area follow a grid of their own.
There are, however, indications of a more sophisticated layout technique at Akrotiri involving, most probably, the use of a measuring system and clear mathematical calculations. This is attested in the large Xest 4 building with its unique ashlar facades of well-dressed tuff; a building that differs in many ways from the typical 'house model' of Akrotiri. On these facades one may observe, for instance, a standard and well-calculated diminution of the height of each course from bottom to top. Moreover, the fact that almost all the examples of mason's marks at Akrotiri derive from this building is probably another indication of the high standards of its architectural design, since these signs seem to be related directly to the issue of design and calculation.
The aspects of architectural design at Akrotiri discussed above have offered, hopefully, an insight into the potentialities of an approach to understanding the architectural relics of our past through a systematic analysis of form and structure. That leaves us with the third component of architectural design - function. Some years ago, the Fourth International Symposium at the Swedish Institute in Athens ended with a despairing question by A. -C. Nordfeldt who asked the audience 'if we can expect a characteristic architectural unit to have one function' (Hagg and Marinatos 1987, 330). Well, architecture alone can not answer questions of function. Though 'form follows function' is a generally accepted statement, it is utopian to ask 'form to lead to function'. Architecture can barely give the general outlines, determining to some extent the range of activities that can take place in a specific space-cell, as well as those excluded. 'The architecture of an age or place is not the passive imprint of thought and action, the direct and transitive "reflection", the mold-negative of systems of thought and value.' (Preziosi 1983, xxii.). So function can be approached only through an interdisciplinary study involving as many aspects as possible. But what the study of architectural design has shown is that before we start naming functions for each building and everyone of its rooms separately, we must try to understand the overall picture of our data and detect the underlying 'models'. Before we occupy ourselves with the variations - and focusing on one variation alone can be very misleading - it is the models that we must analyze, so as to approach as closely as possible the mind of the prehistoric architect, ready to start his work all over again.
Source: "Thera and the Aegean World III, vol I" (pp 44 - 56)
Author: C Palyvou