The Architecture

Conceptual Design

Gastromeni, the land we purchased on Kythnos in 1972, was on a remote bay with no electricity, no readily accessible water, no roads, and no vegetation beyond thorny scrub. It had, however, three attributes that charmed us: it comprised a self-contained cove, assuring us of privacy; it faced west, so the sun would set into the sea each evening; and, most importantly in a country where land is scarce and therefore precious, it was affordable.

Our studies suggested two conclusions for approaching the building of a house that would be affordable, comfortable and sensitive to the island vernacular:

We recognized two alternatives: either find a way to use the stone available on the site in a visually pleasing, economical and structurally sound manner, or adopt a lightweight construction that would minimize transportation cost and reduce the thermomass of the enveloping structure of our house.

Two early schemes served to try out some ideas; their critical evaluation led to the concept that we finally utilized.

Concept A

An architect of Christos' generation cannot start considering the weight of a structure without thinking of R. Buckminster Fuller and geodesic domes. Concept A comprised a three-quarter, 40-foot geodesic sphere, of lightweight material, placed on a drum. Half of the drum was a masonry retaining wall, to be built against an excavated portion of the hillside, while the other half consisted of massive piers radiating from the center and alternating with considerable expanses of glass shaded by the flat ring beam that supported the dome. Internally, the ground level living/dining room was overlooked by a mezzanine-level master bedroom and study.

This concept was theoretical and largely utopian. While it had the advantage of combining locally procured masonry with lightweight imported material, it lacked site specificity – both culturally and topographically. It would require extensive digging and filling of the sharply sloping terrain to create sufficient flat surface to accommodate the design, and the structure could not be expanded without utter disregard for the initial idea. And, it would require a considerable stretch of the imagination to relate the form to Cycladic architecture.

Concept B

This concept explored the use of local stone, mostly schist, through its incorporation in cast concrete walls and piers. The idea was to place large slabs of schist vertically against the interior faces of the formwork, and then fill in with concrete and smaller pieces of stone. This method is called “dumb-walling” (i.e., “any dummy can be a mason”). It offered the following advantages: it reduced the volume of concrete to be mixed, and it eliminated the major weakness of the local schist (water-permeability) by using it as a vertical (face) and random fill material. The design was guided by a tartan grid that enabled the building to climb the terrain with only minor excavation. The idea of a lightweight roof of minimal thermomass and good thermal insulation was further explored in a series of ferrocement vaults.

To explore the potential of this idea, we constructed a short wall using the dumb-walling method. This experiment quickly proved that Concept B was not viable. Although stone is readily available, it is not readily accessible, and would have to be quarried and carried to the building site. Nevertheless, the modularity of the design provided the potential for human scale; an irregular perimeter contributed to the complexity of the form; the grid allowed for future additions; and the use of ferrocement ensured for lightweight construction.

Rejection of the first two concepts led Christos to consider the possibilities within polyhedra forms for building a structure that would incorporate light-weight construction and preserve the two main attributes of the Cycladic vernacular: human scale and complexity.

The schematic design for our buildings evolved from the lessons learned in the first two preliminary concepts. From them, the idea of a lightweight structure of polyhedral form—polyhedra forming the basis for geodesic domes—began to crystalize.

A design for a dwelling based on polyhedra was enticing for several reasons. The geometrical systems present in polyhedra offer the potential for both practical advantages and visual order. Economy of structure, light weight, and speed of erection combine with intriguing spaces and exciting forms.

Structural stability is derived through geometry. Moment connections can be eliminated through triangulation (incorporation of triangles) and, therefore, the amount of material for a structure can be greatly reduced. At the same time, the geometry offers a spatial luxury and excitement that is difficult to parallel with conventional construction under comparable economic constraints.

A polyhedron is a geometrical system comprising three simple elements (vertices, edges and faces) that readily translate into corner joints, edge joints and panels. Yet these simple components result in forms that are visually complex, and in spaces that are dynamic, especially when the polyhedra are truncated and aggregated. The reduction of the building envelope to many small planes lends an inherently human scale to the dwelling, appropriate for house design.

Polyhedral forms, therefore, display the two attributes that provide much of the appeal of the traditional villages on Kythnos, attributes that are usually missing in neo-traditional houses: complexity and human scale.

As the ideas for our house design began to crystallize, this potential for intimate scale and complexity in polyhedral forms became compelling, even more so when the regularity of the geometry was seen against the free form of the terrain. The juxtaposition of regular geometry with irregular form can be viewed as an apt metaphor for the duality of the Greek character (Plato vs. Zorba or, as Patrick Leigh Fermor defines it, the Hellene vs. the Romios) as represented by the classical Greece of cold rationalism and the emotional Greece of Oriental influences. The metaphor extends to the architecture of present-day Athens: the order of the Parthenon rising from the irregular labyrinth of the Plaka.

Christos' research into polyhedra yielded the designs for the Gastromeni buildings

Christos’ extensive research on the architectural potential of polyhedra was based on the capacity of a given polyhedron, or a portion of it (a truncation) to serve as shelter. A 2-meter edge was assumed, and the ratio of floor area to envelope was calculated to gauge the efficiency of each volume.

This research identified the superior capacity of the cube/octahedron family of Archimedean solids for aggregations that would produce efficient shelters. When criteria such as the need for vertical, as opposed to sloping, walls and for combinations that would not introduce odd shapes or spaces were considered, one particular polyhedral solid stood out as best suited for a family dwelling: the small rhombicuboctahedron.

Small Rhombicuboctahedron

At Mikro Horio, the adoption of the small rhombicuboctahedron and associated polyhedral geometry makes sense on several levels:

The small rhombicuboctahedron comprises the eight triangles of the octahedron and the six squares of the cube in pure form. Twelve “filler” squares bring the total number of faces to 26. For architectural use, this polyhedral solid is truncated. That is, nine faces are removed to yield an octagonal floor plan defined by eight vertical squares. Springing from this octagonal drum are four sloping squares and four sloping triangles, with a horizontal square at the top to close the volume.

For the first structure, two truncations of small rhombicuboctahedra were fused together by overlapping two of the squares of the floor plan, to provide the final design for the first phase of the house.

Christos produced drawings and the model for a trial structure constructed of pre-fabricated panels, which would be transported from Athens to the island, and erected on a prepared foundation base.

By 1975, the design was complete and we had the funds and a summer’s worth of time available to put theory into practice.