The Study of Mouldings in Greek Architecture
At the time in 1930 when a study of the changes in the mouldings of classical architecture was begun, little or no attention had been paid to the actual three-dimensional form of those decorative projecting elements which limit and define the structural elements of Greek architecture. When it was obvious that there were indeed striking changes within a given type of moulding in Greek architecture, the question naturally arose as to whether they were accidental or followed some recognizable principle or development. If such a study could be undertaken, it meant, it was felt, that the pieces must themselves be looked at, and looked at as three-dimensional; further, it would take more than the actual autopsy of each piece. Some record of one to be compared accurately with that of others, namely drawings, was needed. That these drawings must therefore be actual size and mechanically as accurate as humanly possible was evident.
The only actual-size drawings of profiles of Greek architecture known to me, at least at the time, were those done of the profiles at Assos by Francis Henry Bacon and published in the volume of the investigations at Assos. Uncle Bacon, as he was affectionately known to us who had the privilege of knowing him, was personally delighted when I met him and he learned what I intended to do, and it should be noted that his interest in the study continued throughout his life. The other person strongly interested from early on was that exquisite draftsman of Greek architecture, including mouldings, Gorham Phillips Stevens, but it was Prentice Duell who first supported my idea of the possibility of real historical significance in looking further at the details already observed and who made possible through his generous backing of the project both the field work and the publication throughout.
It was the director of the American School of Classical Studies at Athens, Rhys Carpenter, who agreed instantly with Prentice Duell's suggestion, and supported and assisted it completely at all stages; it was he who had first introduced so many of us undergraduates at Bryn Mawr College to Greek mouldings. The work began then under the direction of the American School at Athens in 1930. First there was the prerequisite: drawings actual size and mechanically accurate. It was Richard Stillwell, architect of the American School, who said "even negative results would be good at this time; but you need a template such as I have here at Corinth to get these profiles recorded mechanically accurately."
So, with the help of a Maco template the work was begun (Fig. A a-b; also PGM pl. F 14, 15; for a description of the procedure, see below, Appendix A). The plan of action was to visit, examine, and draw every profile from every known excavated site in Greece and on the islands and in Greek cities in Asia Minor. The natural place to begin was a site where there was a range of time and character of buildings and monuments and a few fixed dates, namely Delphi. By the time the work at Delphi was well underway, the general results which the work of later months and years confirmed were already immediately obvious, namely that there was indeed a logical development in the profiles of Greek mouldings chronologically.
From the two fundamental forms available to any builder and extensively used by the Egyptians, namely a convex and a concave curve, known as half round and cavetto, the Greeks developed their series of tvpes (here Fig. B; PGM pl. LXXIX): out of the half round the ovolo, and from it the cyma reversa; out of the cavetto the hawksbeak, the cyma recta, and the scotia. These types have served as the basis of many architectures through time and space ever since. It became clear early in the study that the chronological development of each of the types was universal throughout the Greek lands and this proved to be the case as the study continued to include the Greek mainland, the islands of the Aegean and of the Adriatic and the Greek cities of Asia Minor (including as well the pieces from these sites to be found now in European museums). Although the two groups of Greek peoples, the Dorians and Ionians, developed different types of mouldings (the Ionic the ovolo, cyma reversa, cyma recta, scotia, and the Doric the hawksbeak, cyma recta), the principle of development was the same in both.
The Development of Greek Mouldings
In the development of each of these types the fundamental principle which appears to govern the change or logical development as we see it from the perspective of the whole history is gradual projection: the upper part in the case of the crowning or the lower in the case of the base moulding, i.e., the most projecting part of the material is made to project gradually more and more. Beginning with the half round (here Fig. B; PGM pl. I), the upper half of the top is gradually pulled out so that one moves from a round to an oval at the same time as the point of greatest projection of the whole moves toward that projecting top as an ovolo (here Fig. B; PGM pls. III ff., X, XII-XIII et al.). In this way the light captures the higher point of greatest projection, creating a more subtle effect below. According to the position in the building, all types of mouldings will vary, but none more than the ovolo which serves so many places (PGM pls. XIV-XXIV). The one thing it never does is to become a base moulding.
By at least the middle of the sixth century B.C. the idea of catching the light in yet another effective way caused the very bottom of that ovolo to be given a counter curve, the beginning of a concave, and so begins the history of the cyma reversa (here Fig. B; PGM pl. XXV). Its development is along the lines of a gradual increase in the depth of that concave, and gradually the concave section assumes a larger proportion of the total height of the profile, larger than the convex section; this is true whether it is a crowning (PGM pls. XXVII-XXVIII; XXX-XXXII) or a base moulding (PGM pls. XXXVII-XXXVIII); the projecting portion becomes a lesser part of the total height of the moulding.
As noted above, these two developments of the half round, namely the ovolo and the cyma reversa, are the contribution of the Ionian tradition. It is the Dorians who picked up that other curve, the concave, to create a type to serve all their needs. Beginning with the cavetto (here Fig. B; PGM pl. L) they first cut a tiny nick into the upper fascia portion and balanced it with a tiny nick at the bottom where the concave cut-back begins (PGM pl. LI). Very soon this almost beak-like suggestion is marked off clearly somewhere down along the vertical of the concave portion. Gradually that upper portion from the fascia develops along the lines of an ovolo, the lower portion gradually cutting in further and further. At the same time, the concave is cut up in under the lower point of the beak and the lower demarcation of the curve from the vertical of the block moves upward (PGM pls. LIII-LIV).
Gradually in the course of the beginning of the fourth century B.C. the oval upper projecting part of the beak, which has been following the career of an ovolo, continues to do so to the point of turning a reverse curve at the bottom and so becoming a cyma reversa. The undercut now has become smaller and smaller in relation to the entire profile, and the base line disappears (PGM pl. LV). This hawksbeak becomes the crowning Doric moulding used for all crowning positions (PGM pls. LVI-LXI). A very special version was developed in Athens and used for only a century as the soffit moulding of the raking geison of a pediment (PGM, pl. LII). It is obviously not effectively adapted to a base moulding.
The second development from the cavetto is for the purpose that the cavetto first served in the Egyptian and then Greek architecture, namely the topmost element, the sima, of the building. It was simply the matter of turning a reverse curve, that is a convex curve, at the end of the vertical of the cavetto profile. From the late fifth century B.C. on this cyma recta became the favored crowning profile of a Greek building when stone was used in both Doric and Ionic orders (PGM pls. XLI ff.). It was almost never a base moulding; it remained for the Romans to develop it for that purpose (here Pls. LVIII, LIX).
The Ionian Greeks turned to the concave for their column base needs and developed the scotia on which the half round would rest. The earliest is a slight cut-back through the whole height of the disk on which the half round rests; later as the concave deepens, it is provided with a small vertical fillet at both top and bottom (here Fig. B; PGM pl. LXVI).
It must be remembered that from earliest times the masons were conscious always of the true three-dimensional character of the projecting moulding and at pains to emphasize the actual profile of the solid material by the ornament carved or painted upon its face. The profile of the actual piece forms the basis of the two-dimensional ornament which is carved or painted on it, e.g., a half round is doubled with a half round in the opposite direction and that makes a round, that is the bead which decorates a half round (PGM pl. F). When the half round begins to become an ovolo by being nicked in at the bottom, that profile doubled makes a kind of very early egg. As the ovolo profile changes, so does its egg (PGM pls. A-C, E). When that profile of the ovolo is turned into a reverse curve at the bottom and that line is doubled, it becomes the leaf of the Lesbian leaf which ornaments the cyma reversa (PGM pls. D-E).
To go back to the cavetto. The element of the cavetto is a twofold form from the very beginning. The profile of the concave doubled makes a kind of standing vertical leaf curved at the top, and that is painted into that concave curve at the top, i.e., the Doric leaf (PGM pl. F 8). It must be linked to the vertical portion above so the same tip of the leaf is turned upside down and meets the leaf on the concave below. When that cavetto has a curve at the bottom and the whole profile tends to project more at an angle than stand vertically, it becomes a cyma recta. The question arises as to what ornament that profile suggests. It was the Greek palmette and lotus that suggested themselves; they were used together in alternation. In an alternative form the double curves are suggested by running spirals combined with leaves (PGM pl. F 13).
That brings us back to our consideration of all of these ornaments. In each of them the rhythm of movement along the mouldings needs to be emphasized so that there needs to be an alternation—something to separate one bead or one egg or one leaf from the next so that the rhythm, the movement, along the line is suggested. There is, therefore, always a little point between the two elements; gradually it breaks out between them so that it becomes a dart or reel which separates the egg or the Lesbian leaf. In the cyma recta the lotus and the palmette alternating carry out this principle most effectively.
The Tradition of Greek Mouldings in South Italy and Sicily
A natural next question was what were the Greek colonists in the West doing with their mouldings. Thanks to the support of the American Academy in Rome, a fellowship, strongly endorsed by William B. Dinsmoor who took a special interest in this study, made it possible to try to provide some answers. These were first presented in Profiles of Western Greek Mouldings, volume XIV of the Papers and Monographs of the American Academy, in 1952. The Greek colonies in south Italy and Sicily were pioneers and they exhibited the characteristic qualities of pioneers in any age. It was not surprising then that the observations and recorded drawings of their architectural mouldings reflected three clearcut qualities. First, they carried with them from home the profile types of mouldings which they were accustomed to use in their buildings at home, be it a Doric or an Ionic center. These Western Greeks were themselves following the same general principles of Greeks everywhere as we have seen above, and in addition, they kept in touch with the homeland by trade as well as for other reasons, so that newer followers of the earlier settlers would bring with them the newer developments from home. In other words, the types used in the West were those of the homeland, only they lagged a bit chronologically (here Pls. A-B). A second quality of these Western mouldings came from the pioneer spirit of their carvers: they are "bigger and better." It is thus not surprising that the mouldings are larger in proportion to the architectural member they ornament than they would be for the same member in a building the same size at home (here Pls. A-B). A third quality comes naturally to people who move out to make a new life; they not only get along with each other and the people they find but above all share with each other what they brought from home. It is in the West, therefore, that we find a mixture of Doric and Ionic in the same building well before such a mixture appeared in a building in either Doric or Ionic homelands (here Pls. A-B). The hawksbeak profile in itself is augmented with a sizable base half round in numerous Sicilian temples (e.g. at Selinos). In south Italy, an ovolo is sometimes substituted for the proper tacnia and regulae to crown the epistyle in a Doric building, e.g., at the Lucanian Heraion and, at Paestum, the Basilica and Temple of Demeter (the name still accepted when PWGM was published is retained here for convenience in reference to that volume).
The most striking distinctions of western Greek mouldings from those of old Greece are those of the sima, usually terracotta. Not only are they much higher but they are composed of a variety of profiles, unlike those found elsewhere, including both Doric and Ionic elements and perhaps (probably, I dare to say) inspired also by neighbors on the Italian peninsula.
There is a further distinction between the mouldings of old Greece and the West which was coming clear by the time PWGM was actually published. By that time it had been possible to investigate the mouldings of the non-Greek peoples of the Italian peninsula (PWGM, pl. XXII 8-11 and pl. XXXI 13, 15). The Italic tradition to be discussed below had its impact on Greek forms when Roman domination extended throughout what had been Greek territory.
The Traditions of Mouldings in Ancient Italy
These Greek colonists were not only visiting, sharing, talking with other Doric and Ionic colonists but also with their nearer neighbors with whom they were trading and living happily—the peoples of the northern part of the Italian peninsula, the Etruscans, the other peoples of central Italy, the Romans, the Latins, and the others whose architecture had been so strongly influenced by the Etruscans. As early as the earliest colonists, there was contact with the coast of Italy north of Greek lands, and there may well have been no little sharing of ideas between the colonists in Sicily and the people in Etruscan hilltop communities about how to use terracotta in the upper reaches of buildings, in size and height, if not profile (PWGM pls. I, II, XI, and XII). Clearly we know that by the time the Romans had themselves come in contact with cities in south Italy, in the third century B.C., not only were they strongly affected by the Greek forms they saw, but we recognize that those earlier Etruscan and Roman forms were creeping into the Greek settlements.
This brings us then to the third of the three main areas this study of the profiles of architectural mouldings has included. When the author, waiting at the American Academy in Rome in the Fall of 1936 for the requisite permissions to go south to investigate West Greek architecture, moved around Latium and up into Etruria it was instantly obvious that here was a world of a totally different architectural concept; the idea that Etruscan and Roman architecture was simply a version of Greek architecture was not true. It was clear that one central form was the foundation of Etruscan architectural embellishment not only in Etruria, but it was carried by the Etruscans throughout other Italic lands and to Rome and spread throughout Roman lands as the Romans expanded. In all of the stone portions of their buildings that one form was the convex curve (Fig. C the podium at Ardea). It might vary in proportion from a full half round to a quarter round or a quarter oval, but it was a round developed for whatever the necessity: the base from which a wall or column arose, later the crown of such a wall or altar or small object or column. There are, of course, many differences in the rendering of this round.
It was the American Academy in Rome officers, most of all the then Professor in Charge of the School of Classical Studies, Frank E. Brown, who were very interested in seeing me investigate these Etruscan and Italic forms. They made this further investigation possible by another appointment as fellow in 1949-1950 at the Academy. It was clear that there are many differences between the various sites, but they are emphatically not in general chronological as in Greek lands, though there are some chronological distinctions to be made. The distinctions are primarily geographical in the way in which each separate Etruscan or Latin settlement chose to use that basic round (henceforth called the Etruscan round), either alone or in combination with larger or smaller versions, and occasionally also the concave might appear toward the top of funerary monuments, and, of necessity, in the terracotta revetments used to cover the timber in the upper part of buildings.
The Etruscan round is used individually at different sites, especially for funerary monuments which are very distinctive in each place. For buildings the available stone was used only for the lowest part, whether simple platforms or podia, and for the column bases and capitals. The round serves all needs in stone; its proportions vary with the site, but there is never any ornamental pattern either painted or carved. The superstructure is entirely of timber coated with terracotta plaques; the horizontal beams and the roof, as well as the ends of the projecting rafters must be protected.
We have seen in Greek as in Egyptian architecture that it is the concave rather than the convex that is needed to crown the top of the building where there is a question of protecting the building from the weather; the Etruscans used the natural and obvious concave just as the Greeks and the Egyptians. The Etruscans ornamented this concave by filling it with a row of convex tongues letting only the top of the profile project to throw off the water. This string of three-dimensional elements has nothing to do with the painted leaf used by the Greeks on a cavetto. The Etruscan senses the necessity of emphasizing that three-dimensional concavity of the terracotta crown and changes the decoration into a concave tongue, the top of which curves over and suggests the actual profile of the terracotta. This row of tongues or strigils has no darts between; a continuous row of tongues in an Etruscan sima is very different from the early continuous row of leaves painted on a Greek cavetto which from the beginning has a sense of alternation.
Only when the Roman craftsmen, perhaps either as part of the army or as close followers, came in direct visual contact with the buildings of the Greek settlements of southern Italy did they see, with that universal Roman genius of recognizing and making use of anything good among peoples that they conquered, the possibilities of Greek forms. We have noted that as early as the early sixth century B.C. the Greeks had seen the possibilities of turning a reverse curve at the bottom of the ovolo and creating a cyma reversa (here above, p. xiv; PGM pl. XXV). The Romans immediately saw the value and superiority of this double curve profile over the single big round as a crowning moulding and it began to appear in the third century B.C. as the crown of the podium on which their temples rose (here Pl. XLV ff.), and spread rapidly to other building types. It is interesting to notice that the Romans understood and did not forget the origin of a cyma reversa in a round form: their earliest examples of the cyma reversa and many in succeeding centuries emphasize the convex portion at the very time the Greek forms which the Romans saw already were emphasizing the concave element.
The Romans also recognized the Greek base of an Ionic column as something more effective than their single round to carry the eye and the weight from the vertical column out into the horizontal floor. Of course the Romans thought they could put one round on top of another and let the lower one project. What the Romans failed to recognize was that the concave scotia of the Greek Ionic base itself projected beyond the upper torus, and so the Roman Ionic column base, which has had such a future in all generations ever since, is fundamentally an Etrusco-Roman rather than a Greek column base. Another base form which the Romans developed, again with enormous potential for use alone and in combination with various Greek forms, was the cyma recta. It was created by the turning of a reverse curve at the base of its own round and became the Roman late Republican and Imperial base moulding par excellence for temple podia and other types of monuments (here Pls. LVIII-LIX). The cyma recta was not used as a base moulding in Greek architecture, with one or two rare exceptions, even as the oval was never used as a base moulding in Greek architecture. Thus these two essentially Etrusco-Roman mouldings, the Roman version of the Ionic column base moulding and the cyma recta base moulding, stand in sharp distinction to Greek usage.
Before stating the general conclusions reached from the discussion above, it may be useful to review for purposes of comparison the main distinctions between the Greek and the Etrusco-Roman traditions:
1. Using the fundamental convex half round and concave cavetto, Ionians developed the ovolo from the half round, and the cyma reversa from the ovolo. From the cavetto, Ionians also developed two forms of scotia for bases and the cyma recta for the sima. Dorians developed the hawksbeak from the cavetto for all purposes, and also the cyma recta for the sima.
1. A large convex round (quarter to half round or oval) of varying different proportions for all stone purposes. A small concave cavetto for terracotta revetment crown or sima.
2. Unit of ornament, painted in Doric, carved in Ionic, doubles the line of the profile. Units alternate with point or dart.
2. No ornament on any stone round; terracotta cavetto: solid convex tongues, then concave strigils. No alternation.
3. Development is chronological from the seventh to the second centuries B.C., and universal throughout both Doric and Ionic areas following the same principle everywhere. Exception: south Italy and Sicily lag in time.
3. Different forms of Etruscan round and combinations are geographical, not chronological.
4. Mouldings chiefly crowning and relatively small in proportion to member crowned. Exception: base for Ionic columns. Mouldings larger in south Italy and Sicily, the pioneer spirit.
4. Round mouldings always large in relation to structural elements both in buildings as bases and in other base and crown uses: funerary monuments, altars, columns. This large scale retained when Romans took over Greek cyma reversa for podium.
5. Half round for Ionic column bases and later other bases combined with base cvma reversa and/or base cavetto, but never an ovolo base moulding.
5. Round (quarter, half, full circle or oval) in base form is the Etruscan moulding, used regularly as base moulding in Etruria and wherever Etruscans went.
6. With rare exceptions, the cyma recta is not a base moulding.
6. Romans develop cyma recta as regular base moulding, usually in combination with a fascia or half round below.
On the basis of these specific comparisons, we can deduce that it was the welding of the two traditions, the Greek and the Etrusco-Roman, which formed the basis of the Classical style as it is known today. This welding was the product of the genius of the Romans in the latter centuries of the Republic and was the foundation of the Roman architectural style which developed in the empire and went throughout the civilized world of the West. In the latter years of the empire it disappeared temporarily only to resurface again and again in many times and places throughout the world even into our own day.
Even though the structural inventions of Roman architecture, the arch principle and the concrete material, which made possible the vault and dome, are probably the greatest contribution of the Romans to the history of architecture, the orders remained an ornamental, if not also a structural, essential sheathing of the inner brick-faced concrete structure, and the mouldings are an essential and important part of the orders. Their significance therefore, minor as they may seem in some respects, is indeed major.
Since many of those interested in mouldings for a variety of reasons have requested an account of how the drawings were made, the following technical description is given here:
- Maco Template (A British instrument designed for use in machine shops), largest size available, 26 cm.; staves 14 cm. long and 3 mm. wide. (The standard contour gauge available today is only 6 in. long [ca. 15 cm.; staves 8.5 cm. Iong and ca. 1 mm. wide] ), and therefore not as useful for measuring large mouldings.)
- Carpenter's square, at least 10x15 cm.
- Plumb bob
- Flexible lead wire (electric fuse wire was used)
- Block of drawing paper
- Sharp hard pencil
- Meter stick
- File size note paper
At site, walk the entire area noting every piece with a moulding.
Select the piece of each architectural member whether in situ or on the ground on which the moulding is best preserved.
Clean the block or fragment with excavation knife and/or brush and fingers.
Hold template against best preserved point (Fig. A a-b), adjusting it with the aid of the square so that the vertical central bar of the template is at right angles to the top or bottom surface of the block, or use a plumb bob if the block is in situ.
Push the staves of the template in against the block with a smooth stick and then with the fingers to be sure it is firmly set against the whole height of the profile.
Apply the template to the drawing pad making sure the vertical matches the vertical.
Draw with a pencil against the inside of the template which has taken the profile.
1. When the ornament is carved, note which part of the pattern retains most of the original uncarved profile, and take that profile; sometimes both the center of an egg or leaf and the dart when put together give the most complete profile.
2. When there is an undercut as in a hawksbeak or scotia, draw the full vertical profile as always. Then place a piece of suitable length lead wire from the vertical up against the undercut and down to the vertical of the block the template has recorded. Withdraw wire cautiously and fit it on the template drawing and draw in the undercut from the wire. (It will not have changed significantly if the wire has been fitted accurately to the template drawing.)
3. When moulding is higher than template, more than one template drawing must be made with point of overlapping carefully marked so that the several drawings may be pieced together accurately on tracing paper to give the full profile.
Record overall dimensions of block and details of ornament, including unit measurement.
On note pad record identification, overall and detailed dimensions, and any distinctive details of original work: weathering, damage, re-use.