Acting in conjunction with the downward force is an outward one. The lower parts of a construction tend to spread out as they give way under the weight of the superincumbent masses; if they are very much broader than the latter, they give the impression of great weight carried. As a result, a horizontal line is introduced, and the longer it is in comparison with the vertical line of height, the heavier the effect. Compare, for example, the impression made by a tall and thin triangular shape, with a low and broad one; and compare also the relative lengths of the horizontal and the vertical lines. The former shape seems simply to rise, while the latter lifts. We seem to observe the working of this outward force, as Lipps has remarked, in the spreading out of the trunks of trees at the base and in the feet of animals; and we feel it in ourselves whenever we spread our limbs apart to brace ourselves to withstand a load.
Whenever the outward force is resisted, it gives evidence of the existence of a force operating in the opposed direction—inward. Without this force, the lower parts of a construction would lack all solidity and spread like a molten mass on the ground. This is especially striking where the material, instead of spreading outward and downward, seems to press itself inward and upward. Compare, for example, a shape whose base-line is smaller than the line of its top with one in which the reverse holds true. The former gives the impression of lightness and agility, with a prevailing upward trend, the other an impression of weight and heaviness, with a prevailing trend towards the ground. Obviously, the outward and the inward forces are correlative and complementary: we have already observed that a construction would collapse without the inward; we can now see that it would disappear entirely without the outward. Obviously, also, the inward and upward go together, and the downward and outward.
Even a plain rectangular wall manifests the interplay of these forces. The horizontal dimension represents the downward and outward force of the weight; the vertical dimension, the upward forces, which prevent the wall from collapsing in itself and hold it upright; while the lateral boundaries give evidence of the inward tension that keeps the mass together. But the most beautiful expressions of architectural forces are to be found in the historical styles. In each style there is a characteristic relationship between the forces, imparting a distinctive feeling. I shall offer a brief analysis of some of these.
Many have recognized that the classical Greek construction, as illustrated in the Doric temple, expresses a fine equilibrium between the upward and the downward forces, embodied in the vertical and horizontal lines respectively. The upward force is manifest primarily in the vertical columns, and is emphasized there by the flutings, the slight progressive narrowing toward the top, and the inward effort of the necking just below the echinus. The downward force is embodied in the horizontal lines of the lintel, architrave, cornice, and in the hanging mutules and gutta. The two forces come to rest in the abaci, which, as the crowning members of the columns, directly carry the weight of the entire entablature. The equilibrium between the horizontal and the vertical tendencies is, however, not a static but a moving one; for the two opposing forces are present in every part of the building from the stylobate to the ridge of the triangular pediment. The downward force is already manifest in the widened base of the column, where it works in conjunction with the inward tendency, and shows its effect at the critical points at the top of the supporting column—in the spreading echinus with its horizontal bands beneath and in the horizontal lines of the abaci. The upward force, on the other hand, is continued right through the solid mass of the entablature, in the vertical lines of the triglyphs, in the antefixes, and even to the very apex of the building, where the ascending lines of the triangular pediment meet. The resulting total effect is that of a perfect, yet swaying balance.
The aesthetic effect derived from the interplay of forces in the Ionic form is similar to that in the Doric, only more delicate and elastic. The slender columns, being less rugged and resistant than the Doric, seem to transmit the weight supported, which shows itself, therefore, in the outward spreading molded base; but this apparent lack of strength in the column is compensated for by the elastic energy in the coiled spring of the volutes, upon which, with the slight mediation of a narrow band, the entablature rests. Here most of the upward energy of the Ionic form is concentrated; for although the dentils of the frieze perform the function of the triglyphs, they are too small to do it effectively; the style lacks, therefore, the gentle harmonizing of forces all over, characteristic of the Doric, and evinces instead a clean-cut elastic tension at a given point. This effect is, however, somewhat softened by the breaking up of the downward force of weight by means of the recessed divisions of the architrave. In the Corinthian capital, which has the same general feeling as the Ionic, the elastic tension is still further diminished through the renewed emphasis on the mediating abacus, the reduction of the size of the volutes, and the overhanging floral carvings. However, by reason of the strength given by the bell and the projecting outward and upward curving form of the abacus, the suggestion of weakness in the Corinthian form is overcome, but the gentleness remains.
If the Greek construction expresses a balance between the upward and downward forces, the arched forms that followed express the victory of the upward. In the arch the upward force, instead of being arrested where the support meets the mass to be carried, is continued throughout the mass itself. Of the two chief types of arches, the round and the pointed, each has a specific feeling. We shall study the round form first, where the vertical tendency is indeed victorious, but only through reconciliation and compromise.
In the round arch all four forces are beautifully expressed. The upward is manifest, first, in the vertical pier, which acts very much as the column does, and, in Roman work, was often replaced by the column. The opposing downward force is expressed in the horizontal upper bound of the arch and in the line of the impost, also horizontal, which breaks the vertical line and so marks the place where the two forces come into sharpest conflict. In this conflict, the vertical is victorious; for, instead of being stopped by the impost, it is carried up throughout the entire construction by means of the upward and inward curving of the arch. The very curve of the arch shows, however, that the victory is not absolute; for its circular form is obviously determined as a compromise between an inward centripetal force, moving upward and diminishing the breadth of the arch to a mere point at its apex, and an outward centrifugal force, gradually spreading the arch downward until it reaches its greatest breadth at the impost, where it is arrested by the opposing vertical force in the pier. To the historical imagination, the round arch seems, therefore, to express the genial classical idea of a control by the higher nature which nevertheless did no violence to the demands of the lower. In the spherical dome the effect is the same, only the interplay of forces operates in three dimensions instead of two.
When arches are superposed, the upward movement proceeds in stages, beginning anew at each horizontal division of the wall space. The use of entablatures applied to the wall and of engaged columns, common in Roman work, seems to involve an attempt at a fusion of two contradictory styles, and is usually condemned as such. This contradiction can be solved, however, by viewing the entablatures as mere weightless lines of division of the wall, usually marking off the different stories, and by viewing the columns in a similar fashion as having no supporting function—which is actually the case—and as simply serving the purpose of framing the arches. At most they merely indicate the direction of the chief contending forces,—the parallel lintels signalizing the force of weight, and the vertical columns, standing one upon the other, pointing the movement of the upward force. They have, therefore, a pictorial rather than a dynamic significance.
Differences of feeling in arched forms depend upon the relative height of arches and supporting piers and columns. The vertical effect is strongly emphasized when the latter are relatively high, while the effect of weight is increased in flattened arches, which for this reason are especially appropriate for crypts and prison entrances. Interesting complications are introduced in arcades or intersecting vaults, where a single column serves as a support for two or more arches; for there the vertical force is divided, flowing in different directions in the little triangular piece of wall between, or along the ribs of the vaults. Something similar occurs in the Byzantine dome on pendentives, only instead of supporting the horizontal weight of a gallery or a vault, the triangular pendentives meet the outward thrust of a superposed dome.
In Renaissance architecture and the modern classical revivals, where Greek and Roman styles are freely adapted to novel modes of life and purpose, no essentially new form was added to architectural speech. There were combinations of old forms into more complex structures, but no new important elements. The most outstanding novelty is perhaps the reversed relation between the whole and the parts. [Footnote: See P. Frankl, Die Entwicklungsphasen der neueren Baukunst, 1914.] In the classic styles, whether arched or Greek, the whole is built up of the parts additively; each is a relatively independent center of energy complete in itself; first the columns, then the architrave, frieze, and cornice, then the pediment; or first one row of arches, then another row on top of this, and so on. Coordination is the governing principle. But in the modern adaptations, even where coordination rather than subordination rules in the pictorial sphere, the whole is first dynamically and the parts are secondary. In the typical Renaissance facade, for example, the arches of the windows are rather openings in the walls than supporting members. They are centers of little eddies of force, rather than independent parts of the main determining stream of energy. The wall rises as a whole to its heavy overhanging cornice, despite the horizontal divisions marking the stories. There are, however, important differences between the various modern types; the earlier Renaissance forms, for example, keeping closer to the antique than the later Baroque and Rococo.