The American Indian/Chapter 6

CHAPTER VI

ARCHITECTURE

The only regions in which building rises to the level of architecture are those occupied by the higher cultures of Mexico and Peru. Roughly considered, there are indications of three centers of development: Maya, Nahua and Inca, though in last analysis we may find but two, the Maya and Inca, using those terms in their broadest sense. The chief characteristics common to both are rectangular groundplans, massive masonry walls, often of rubble, and the absence of the arch. The last is probably the most important factor, for the clumsy method of a stepped ceiling, closed by a slab of stone, not only doomed the builders to narrow rooms, but required very thick, firm walls for their support. The published plans of the most typical ruins show long, narrow rooms or tiers of rooms, the widest so far reported being 14 feet.^[1] That these ancient builders were aware that at best this method of vaulting gave but weak support, is shown by the tendency to support upper stories upon a solid masonry core around which the lower rooms were grouped.^[2] This may also be one cause for the relative infrequency of storied structures and the almost universal practice of securing height and elevation by building upon artificial or natural mounds (Fig. 52). The necessity for narrow rooms no doubt led to the enclosed rectangular court plan, which prevailed both in the North and the South (Fig. 51). Curved or circular walls are very rare and when found are isolated and not a part of a rectangular building. Consequently, we have a plain rectangular contour as a universal character.

Windows are very rare, especially in the North, and the doors are usually rectangular with straight lintels. Colonnades formed with rectangular stone supports are common, and in a few cases we meet with the cylindrical pillar, but it did not develop far enough to constitute an architectural feature. Some remarkable feats of masonry are found, and the skill of these ancients in handling cement and transporting huge masses of stone excites our admiration.

Fig. 50. A Cross-Section of the Temple of the Cross, Palenque, Chiapas.
Holmes, 1895−1897. I

In the essential characteristics we have noted, there is little to distinguish between the buildings of Peru and Yucatan, though, as we shall see presently, they did have important differences. The present state of our knowledge suggests that all the ruins of southern Mexico and of the adjoining Central American states are historically related; but the type seems to disappear toward the Isthmus and reappears in Ecuador. The disconnection of these two centers in the face of their common structural characteristics presents an interesting problem as to how much of this is due to borrowing. If we

Fig. 51. Groundplan of the House of the Nuns, Uxmal, Yucatan.
Morgan, 1881. I

take a little broader view, we shall find certain more widely distributed building concepts. First, the pyramidal mound for burial seems to extend from northern Mexico to the Isthmus and then to recur in Colombia, passing through Ecuador and down into the coast of Old Peru. At least in one part of the Inca domain we find buildings upon them. In fact, their general absence in Old Peru is accounted for by the rocky nature of the country, which affords sites of natural elevation to which buildings were frequently adjusted by terraces.^[3] It may be of interest to note that the pyramid mound both for burial and building sites extends up into the Mississippi

Fig. 52. Elevations and Ground-plans of the Ruin Known as Santa Rosa Xlabpak, Yucatan. Spinden, 1913. I

Valley as far as the famous Cahokia of Illinois, and that this distribution is continuous with the general mound area of the upper valley. In other words, the occurrence of mounds of this type has a generally continuous distribution from the Great Lakes of the North to the coast of Old Peru of the South. Throughout, they are most numerous in level districts.

The northern limits of building attributed to the Nahua are on the Gulf Coast about the 24° of latitude, or in striking distance of the Rio Grande. Though all the later northern buildings are far less preserved than those of the older Maya, they seem to have one suggestive difference, the absence of the vaulted ceiling and the consequent increased size of rooms. The rooms were probably flat and supported by beams resting upon internal pillars where necessary (Fig. 53). It seems strange that the Maya did not make more use of wood, but the Nahua style reminds us of Pueblo architecture, where beams of wood support the ceilings and roofs.^[4] Thus again, we have an interesting case of continuous distribution. It is certain that the large and imposing ruins did not house the bulk of the population. The surviving examples show that the prevailing habitation was a small, rectangular one-room house whose essential structure, when of stone or adobe, was the same as found in the several units of the so-called palaces and temples, except that the roof was thatched. In Peru, the roofs were often supported by ridge poles which would give us about the same interior effect as the stepped ceilings. The walls of the houses take three forms, all of which may be encountered on either continent; namely, stone, adobe and mud reinforced with canes or wattling. Studies among the Pueblos of New Mexico have indicated that when we know more of that area we shall find a period of single detached adobe and stone rectangular houses preceding the composite pile of the modern pueblo. In fact, the Pueblo Indians of the present show a disposition to revert to the detached house, which does not materially differ from a single unit in the village structure. In like manner, we find in Peru a grouping of single houses around a court so as to form a complete enclosure, and the groundplan of these is not essentially different from those of the preceding structures. Similar conditions have been reported for the Maya district.

We see, then, that in at least two particulars we have a broad cultural base for the highly specialized building arts of the Maya and Inca. That all these widely distributed characters result by diffusion from these two centers is scarcely logical, for even cultures are not built of nothing, but all have a long train of historical antecedents. It is much more reasonable to assume that diffusion, and perhaps other factors, brought a certain extended uniformity in house-building before the final burst of higher culture in these two centers. Granting that in this burst they may have been independent, they nevertheless had the same heritage from which to fashion their art.

One argument for their independence is to be found in the secondary decorative features. In this respect the northern buildings are far in the lead. The embellishment of the façades is often intricate and full round sculptures are let into the walls by tenons; stucco reliefs are built out upon rough skeletons of stone work; and elaborate mosaics of separately carved stones are arranged so as to make grotesque faces, as

Fig. 53. Restored Section of the Hall of the Six Columns, Mitla, Mexico. Holmes, 1895−1897. I

well as geometric patterns. A special feature is the use of monolithic monuments commonly called stelæ, placed around and among buildings, the surfaces of which are richly carved with pictographs and hieroglyphs. The exterior and inside ornamentation was often painted in a very skilful way.

When we turn to Peru, such monuments are conspicuously absent and the exteriors of the buildings are, in the main, plain.^[5] Still, we have an approach to it in the celebrated stucco walls of Chanchan, bearing an elaborate textile design, and in the inland we find traces of painting upon smooth stucco, suggesting that there was a great deal of such ornamentation that has disappeared. Then we have a few noted monoliths, as the Chavin stone and the very remarkable gateway at Tiahuanaco. To these may be added the curious sculptures at San Augustine, Colombia. Both the stucco and the monolithic carvings have a certain general resemblance to those of the Maya, but on the other hand, they have great differences. It is also noticeable that they have their

Fig. 54. Reconstruction of Pueblo Bonito, Chaco Canyon, New Mexico. Morgan, 1881. I

counterparts in the textile and ceramic art of their respective localities. Yet, the distinction remains, that secondary embellishment, or what is often considered true architecture, is characteristic only of the Maya type.

Associated with the foregoing culture were no less worthy feats of highway and drainage construction, particularly by the two great military cultures, the Nahua and Inca. In Peru, roads were paved and graded and brooks spanned by stone culverts, many of which are still in use.^[6] These were necessarily formed by huge stone slabs supported by piers. Chasms were bridged by true suspension bridges and in some cases crossed in chairs running on cables. Even a kind of pontoon bridge was in use. In Mexico, the country was less rugged, but the roads were excellent. In both regions the irrigation and aqueduct systems are famous. As all travel was by foot, and only in Peru were pack animals used, these road builders had a somewhat different problem than confronted the users of carts in the Old World.

As we proceed southward in Peru, architecture rapidly deteriorates, disappearing altogether at the River Maule. Thence toward the east in the neighborhood of the Calchaqui, we find rough stone structures of many rooms, not unlike one-story pueblos. Burial is now in urns without true mounds, but many small carved monoliths have come to notice. Once out into the guanaco area we find the simplest kind of skin tent, which in the far south becomes merely a windbreak (Fig. 72). Throughout the Amazon, on the north coast and southward as far as the Suyas, hammocks are in general use and the houses are frequently primitive. On the other hand, very large thatched structures are found, under which, as under one great shed, lives the whole community. So far, there seems to be no consistent distribution of varieties of this type, some are oval and well thatched, some square, and some mere roof shelters. In fact, the only thing essential is a hammock to keep one off the ground and a roof overhead. The whole population is rather nomadic. As we go eastward through the highlands of Venezuela, the court structures of Colombia disappear, but still the prevailing form is the rectangular hut; but in Guiana we begin to encounter the oval thatched house of Brazil. Of some interest are the pile-dwellings of the north coast, now almost extinct, though a few survive in swamps and even on dry land. In some of the inundated districts floating houses are found. Finally, the meager archæological data we have reveal only one important site at the mouth of the Amazon where mound structures have been reported.

The structure of habitations in the United States and Canada has been carefully studied so that we can make very definite statements as to the types and their distributions.^[7] Nowhere outside of the frontier to the Pueblo area do we find buildings of stone until we reach the Eskimo. Consequently, there is very little content to the archæology of architecture, our data being almost exclusively from the surviving tribes. The only building that reminds us of the traits we have discussed in our consideration of the area of intense maize culture was found in the lower Mississippi Valley, a rectangular house with walls of clay reinforced by wattling. Sometimes, as in Arkansas, there were two or three rooms suggesting the houses of Colombia, but these were not the prevailing type on the lower Mississippi.

The Gulf States form a fairly distinct house area. Particularly on the Atlantic side were curious oblong rectangular houses with curved, or bowed, roofs. Their construction was simple, a framework of light poles, lashed into place, with coverings of bark or thatch. (For type illustrations see the Handbook of American Indians.) In the Florida swamps a kind of platform pile-dwelling is found, with roof and open sides reminding one of Guiana types. A very widely distributed structure is an oval dome-shaped house, plastered over with mud, with no opening except the door. In fact, none of these southern houses seems to have been provided with smoke holes, most of the cooking being done out-of-doors.

In northeastern United States the prevailing form among the Algonkin tribes was a low, oval framework of poles covered with bark, mats, or thatch, according to the season and locality. The Iroquois of New York, who are generally regarded as of southern origin, lived in long, rectangular, bark-covered communal houses known in local literature as the "long houses."^[8] The structural similarity of this to one of the southern types is obvious. West of Lake Michigan the dome-shaped Algonkin house often gave way to a rectangular one with a flat roof, and among the Eastern Dakota we meet with this form made by setting up rows of posts in the ground. A little farther west on the Missouri we have what is usually called an earth-lodge, a circular, conical-roofed framework covered with thatch and finally with turf.^[9] However, its distribution is restricted in the main to Caddoan and Siouan tribes of bison hunters, who also raised some maize.

Next, we have a well-known type of shelter to which the Dakota name, tipi, is usually applied. In the East, it appears in northern New England, extending up into Labrador, thence eastward through the great Cree and Ojibway range, well up into the Canadian Northwest. Also, it sweeps down into the bison area, reaching some of the nomadic peoples of the Pueblo area and again invading the salmon area in Oregon and Washington. The other forms of shelter we have noted have all but disappeared, while the tipi is still used by the surviving tribes of this great area. These conditions tend to make it the most typical Indian shelter, and it now has so firm a place in the popular mind that it is used in art and story, regardless of the locality. Not infrequently we see pictures of Pocahontas, Henry Hudson on Manhattan, and even of California incidents associated with tipis, a form of shelter entirely inappropriate. The term wigwam in Colonial literature is the Algonkin name for the oval bark-covered house we have described, and the modern tendency to apply the same name to the tipi has led to great confusion.

It is not to be expected that we shall find a single type of tipi prevailing throughout. The essential structural concept is a tripod of poles, supporting other poles forming a cone.^[10] The base tripod is formed by binding together three or four poles, but in far western Canada these poles sometimes have interlocking forks, a feature also noted in southern Nevada and in the older type of Navajo hogan. Where birch trees grow, the cover is birchbark; in the bison area it is skins. The Ojibway, however, often used mats, as was sometimes the case on the Columbia River. In the far North, we find a pointed skin tent, even forming a summer dwelling for the Eskimo.

For the details of varieties of tipi and their distribution we must refer the reader to the special literature. We note that it seems to follow the outlines of the caribou and bison-hunting areas and is everywhere definitely associated with a nomadic hunting life, for many tribes on the borders used it only when on hunting trips. Its origin and development, therefore, is one of the important problems in our subject and must receive close attention in the future. Curiously enough, the tipi is found in Siberia and has analogous forms in northern Europe, suggesting the possibility of its definite association with reindeer culture.

We have now covered the whole of the northern continent except the western part and the Arctic. The most distinctive structures here are the wooden totem-pole houses of the North Pacific Coast, reaching their highest development among the Haida and Tlingit.^[11] The structural plan consists of four massive, upright timbers supporting two long, equally heavy beams. These are placed parallel about four feet apart and are essentially ridge poles. Around these a rectangular enclosure is made by setting split planks upon end. The ends are gabled and the roof of planks. The only framework is the massive central support, in contrast to which the remainder of the building appears flimsy in the extreme. But we find one feature not so far observed north of the Nahua area, namely, architectural embellishment. The four interior posts are carved in high relief, and outside is the famous totem pole. Paint is used to reinforce the carving, and in addition the front of the house is decorated with one of those curious spread-out animal forms we have noted in the preceding chapter. Had these people carved in stone instead of wood, we should now find their country one of our richest archæological fields, but the perishable nature of their building material has left no records of their past history.

The influence of this type of architecture reaches northern California, for though the heavy carved timbers have a central distribution only, we find the rectangular house of upright planks, with a circular door throughout the coastal belt of Oregon and Washington. In Canada it invades the mountainous area of the Déné, but in Alaska it stops rather suddenly.

Central and southern California present simple but various forms of shelter.^[12] Yet, they may be characterized as shelters of brush and tule reeds. More permanent houses are sometimes formed by setting up poles over slight excavations. Toward the interior we meet with the great Shoshoni range, the characteristic shelter of which is a simple brush-covered lodge. Two forms occur, the precise distribution of which is not yet known, but the prevailing one seems to be a low dome-shaped, grass-covered affair still encountered among the Comanche and the Apache. The other type we have mentioned is a pointed brush shelter upon a tripod of forked poles, a form closely allied to the Navajo hogan and perhaps to the tipi.

Strictly considered, none of these houses so far described can be classed as underground. Yet, some approach this qualification in that they have sunken floors. Thus, in California, the house is often over a shallow pit, and elsewhere it was common to remove the surface soil to expose the clay or other hard layer, loam being too powdery when dry to make a good floor. However, when we turn to the inland Salish tribes of British Columbia, we meet a more distinctly underground house entered through the smoke-hole at the center by a stepped ladder.^[13] The distribution of this form centers very closely among the inland Salish who may, therefore, be considered its originators. The next place where we encounter a subterranean house is among the Eskimo of Alaska. In this case we have two ways of entering: through the smoke-hole, and by a long covered trench, each used according to the season. The Eskimo house, however, is often set over a very shallow excavation and earth heaped over its framework like the earth-lodge of the Missouri. This form of house extends eastward beyond the mouth of the Mackenzie, but from there on timber is too scarce. Stone houses were noted by Stefánsson near Coronation Gulf, and their distribution from that point eastward seems to be continuous. Their roofs are usually of skins, often supported by whale ribs. The snowhouse we all know so well is universal from east to west as a temporary residence, which in summer gives way to a small skin tent. Its long, low, tunnel-like entrance and internal arrangement is the same as that for the earth-covered type of Alaska, and both together may be regarded as revealing the characteristic Eskimo house concept.

Jochelson^[14] has brought together some data for a historical connection between the earth-covered houses of the Old and New Worlds. While it is clear that examples of such dwellings are found intermittently from Europe, across Asia, to America, we do not find the definite structural parallels necessary to form satisfactory conclusions regarding their historical relations. Archæological work has brought to light a somewhat more extensive distribution of such houses in America. Numerous depressions in the upper half of the Ohio Valley have been regarded as old house sites and recently Sterns^[15] located rectangular house pits in Nebraska, but, except in the last case, our knowledge is not definite, and the very perishable nature of the structures so far observed makes further discovery extremely difficult.

In conclusion, attention may be called to one peculiarity of aboriginal house construction. The chimney was unknown. Not even the skilled architects of Mexico and Peru seemed to have hit upon the idea. It is true that in the historic pueblos they are found, but this is generally attributed to Spanish influence. In the older type of pueblo structure only the rooms having open roofs were used as living quarters. Hence, the universal American way of heating a house is by an open hearth at the center with a hole in the roof immediately above.

Fortifications may also be considered under the general head of architecture. At the time of discovery the native villages in the southern half of the eastern maize area were circled by palisades. In the north, the Iroquois possessed such fortified towns and even in New England they were known.^[16] It is now considered that certain rings of earth in New York State mark the sites of palisaded villages, and there is reason to believe that similar redoubts in the Ohio Valley had a like origin. The palisade was used as far up the Missouri as the Mandan villages. In fact, the distribution of palisaded villages is about coincident with maize culture in the East. The only other place for which palisades are reported is the North Pacific Coast, though the usual form in that region was a high rock with overhanging platforms like a blockhouse. Nowhere else do we find fortifications until we reach the Pueblo area. It is true that a number of earthworks are designated as forts, but their use as such is largely hypothetical. Perhaps the best-known example is Fort Ancient in Ohio. In the Pueblo region the houses were so placed, either in cliff recesses, upon mesas, or piled upon each other in such manner as to make other defensive works unnecessary.

In the Antilles and eastern South America the palisaded village was known, but we have no record of other kinds of defensive works. It is, however, important to note that we have here a continuity of at least one trait for the eastern halves of both continents.

Naturally the great military empires of Mexico and Peru developed fortifications. In the former, the road from Tlaxcala to Mexico City was defended by a stone wall about six miles long, faced by a ditch.^[17] The internal citadel of the defense works at Mexico City was about the temple of Tenochtitlan, surrounded by a wall six feet high, where the last stand against Cortez was made. Strange to say, the great ruined cities of the Maya show no definite fortifications. It is, however, in the Inca territory that the greatest systems of defense appear.^[18] Important points on roads were guarded by blockhouses, cities were defended by systems of outlying forts, etc. The most famous fortresses are Ollantaitambo and Sacsahuaman. The latter is distinguished for its remarkable masonry and the former for its internal passages cut in solid rock.

---

1. Spinden, 1913. I.
2. Holmes, 1895−1897. I.
3. Joyce, 1912. I.
4. Holmes, 1895−1897. I.
5. Joyce, 1912. I.
6. Markham, 1910. I.
7. Morgan, 1881. I
8. Morgan, 1881. I.
9. Fletcher and La Flesche, 1911. I; Spinden and Will, 1906. I.
10. Wissler, 1910. I.
11. Emmons, 1916. I.
12. Kroeber, 1904. I.
13. Teit, 1900. I.
14. Jochelson, 1908. I.
15. Sterns, 1915. I.
16. Willoughby, 1906. I.
17. Joyce, 1914. I.
18. Markham, 1910. I.