NATIONAL PARK SERVICE Ruins Stabilization in the Southwestern United States Publications in Archeology 10

The problem of stabilizing and maintaining historic sites is one that has almost as many ramifications as there are individual structures. Since this book originated from experiences derived in the Southwestern United States, including Arizona, New Mexico, Colorado, Utah, southeastern California, and western Texas, it is concerned largely with pre-Columbian masonry structures, post-1540 A.D. adobe structures, and, to a lesser extent, with monolithic soil structures, sometimes termed pise or rammed earth. In the following discussion of Southwestern archeological sites, we will deal first with the problems of structural faults and inherent weaknesses. We will then look briefly into the forces that have contributed to their deterioration. Some understanding of these faults and forces is necessary before they can be effectively counteracted.

Anyone who sets out to stabilize a prehistoric structure will find that he has two conflicting objectives. The first of these is the requirement of authenticity in preserving irregular and relatively primitive construction. Equally important is the need for permanence, the requirement of any sound building practice.

STRUCTURAL FAULTS

The prehistoric mason was a progressive architect and builder for his day. The Anasazi progressed in a short span of 300 years—10 or 12 generations—from pit houses to multistoried buildings of 400 and 500 rooms. In each period and area, he followed a rather set pattern of plan and construction circumscribed by the limitation of local materials. At any one time a village consisted of an accretion of ruined, decaying, remodeled, and new units. Rooms were often built on the partially razed wall of older structures and on loose fill. Foundations were often narrower than the walls they supported. Walls were rarely bonded at corners or other junctions.

The materials encountered in almost all possible combinations include abundant soil mortar with spalls stuck in the surface, soil mortar with unshaped or slightly shaped stone, facing at stone-to-stone contact, dry masonry, wattle-work, and jacal or pole and mud construction. Soil structures were always laid in a plastic condition—in thick courses at Casa Grande, in plastic loaves in turtleback walls of Pueblo I sites, and in roughly squared plastic blocks in ruins on the Animas and at some late prehistoric Rio Grande pueblos.

The following list summarizes the structural faults most commonly encountered:

1. No foundations (fig. 1).

2. Foundations narrower than the walls they support (fig. 2).

3. Construction over loose and unconsolidated fill.

4. Long spans over openings supported by extremely small lintels.

5. Lack of bond at wall junctures.

6. Lack of headers or ties through walls where there is more than one width of stone.

7. The incorporation of large horizontal wood beams in masonry walls (fig. 3).

8. The inclusion in walls, at ceiling height, of horizontal areas of bark, splints, rods, and other ceiling materials.

FIGURE 1. Settling caused by lack of foundations and unstable fill. Bond between stone and soil mortar is poor. Holes are caused by decay or theft of wood members. (Antelope House Ruin, Canyon de Chelly National Monument, Ariz.)

FIGURE 2. Aboriginal construction often left narrow strips at ceiling line. This wall is much wider at the top than in the center. (Room 27, north wall, Pueblo del Arroyo Ruin, Chaco Canyon National Monument, N. M.)

FIGURE 3. The inclusion of heavy timbers within cored-type walls was a usual practice of prehistoric builders. Collapse of walls from flood damage at Chettro Kettle exposed 300 logs similar to those above. (Chettro Kettle Ruin, Chaco Canyon National Monument, N. M.)

A number of post-1540 A.D. buildings at Pecos, Tumacacori, Fort Davis, Fort Union, Fort Bowie, and the early 20th-Century Spanish-American residences at Big Bend National Park, were constructed of adobe bricks. Where either the adobe or lime mortar plaster had weathered from wall core surfaces, and the roofs were either removed or deteriorated, ruins range in appearance from no vestiges at all, through low mounds, to imposing wall remnants rising in excess of 30 feet above present ground surface. While structural faults were not as common as in the pre-Columbian structures (the result of advanced methods of construction) a number of factors, not the least of which was the type of building material, resulted in more rapid deterioration from causes mentioned below.

WEATHERING

The breakdown of an abandoned structure follows a rather set pattern. First, wind- and water-borne material is deposited on the floor. This deposition continues until it is covered by collapse of roofing and dislodged sections of wall. Once the roofing is gone, stability of walls is threatened, and the generally poor materials and faulty construction become particularly vulnerable to the elements. Moisture from snow and rain melt out the mud mortar; thin unbonded facing is separated by frost action; rotting of wood parts such as beams, lintels. or ceiling inclusions results in unsupported overlying masonry; pressure from fallen material dislocates remaining walls, and the accumulated debris ponds surface water which, in turn, subjects basal areas to moisture penetration and erosion.

Depending primarily on the type of construction and location of the structure, these forces proceed at different rates. In small open units which are usually thin-walled structures, the downward erosion is most rapid, and large areas of collapsed wall are rarely found. Within large sites of heavy construction and in protected sites this downward erosion is less rapid. Collapsed sections of large wall areas are more often encountered, the result of horizontal erosion at base and ceiling lines, and the decay of wood members.

The action of natural forces outside the structure also has an important bearing on preservation. Weathering of wall bases and soft sandstone outcrops on which pueblo ruins stand has become serious at Hovenweep and in sections of Wupatki. Important sites in caves are deteriorating at Navajo and Canyon de Chelly National Monuments, the result of wind and water having removed supporting fill and portions of bedrock. Within historic times, about 75 percent of one 40-room site in Chaco Canyon has been lost to arroyo-cutting. As opposed to erosion, the deposition of wind- and water-borne materials adjacent to sites can sufficiently alter the topography to destroy natural drainage and cause flooding.

Walls of pise construction, such as those at Casa Grande, and adobe brick construction at Tumacacori, present special problems. Natural weathering is particularly severe where the protective roofing is gone and the foundation is not waterproof. Erosion proceeds from all sides, especially from the tops and bottoms of the walls. Starting at the bottom of the wall, soil moisture rises by capillary action, carrying soluble salts which are deposited at the evaporation line just above existing ground level. A large proportion of these salts are hygroscopic, and are composed of chlorides, carbonates, and nitrates of calcium and magnesium. This concentration of salts acts in two ways to destroy the character of the adobe at the evaporation line. Crystallization shatters the adobe, causing continual shedding of surface particles. Also, concentration of salts above an unknown limit results in chemical destruction of the cohesive elements of the material and in its further disintegration. (Adobe workers are well aware of this process and do not use soils from alkali bearing areas for their bricks. Furthermore, they dislike the use of water with a strong salt content.) As this concentration increases, the salts, as hygroscopic agents, attract moisture from the air, especially during periods of high humidity, and the process of erosion is accelerated. Erosion of this type results in troughing which, in time, may cut entirely through the wall.

Erosion at the tops and sides of an adobe wall is caused by rain, hail, wind and melting snow which result in vertical rivulets, fissures, and a general thinning and eventual melting of the adobe.

HUMAN AND ANIMAL DISTURBANCE

Unprotected sites have been vandalized for centuries. Such activity is more destructive of prehistoric buildings than is generally supposed. In the first place, the prehistoric inhabitants periodically tore down old houses to reuse the timbers and building stone. Neil Judd makes note of prehistoric vandalism at Pueblo Bonito, at a time following its abandonment, by groups seeking precious objects of turquoise, jet, and shell (Judd, 1964). Similar plundering and salvaging continued into the historic period. Early settlers in the Chaco speak of the Navajo removing wagonloads of timbers from Chettro Kettle for firewood. These same settlers did so themselves, and also used stone from the Chaco Canyon ruins.

Adolph Bandelier, the famous Swiss anthropologist, undertook his first important field project in American archeology—an exploration of the Southwest, in August of 1880. He began with a preliminary study of Pecos. In his report (1883, p. 42) there appears this shocking account of the 18th-century mission church of Pecos:

Mrs. Kozlowski (wife of a Polish gentleman, living two miles south on the arroyo) informed me that in 1858, when she came to her present home with her husband, the roof of the [adobe brick] church was still in existence. Her husband tore it down, and used it for building outhouses; he also attempted to dig out the cornerstone, but failed. In general the vandalism committed in this venerable relic of antiquity defies all description...all the beams of the old structure are quaintly...carved...much scroll work terminating them. Most of this was taken away, chipped into uncouth boxes, and sold, to be scattered everywhere. Not content with this, treasure hunters...have recklessly and ruthlessly disturbed the abodes of the dead.

Prior to acquisition by the National Park Service, a number of military posts either suffered severe depredations or were fortuitously preserved, at least in part, following abandonment by the Army. Fort Bowie was turned over to the Department of the Interior in 1894 and the land was auctioned off to ranchers. Subsequently, local residents cannibalized the buildings for construction materials. Erosion quickly set in, and the post soon fell into ruins. Some of the major buildings at Fort Davis and Fort Larned were occupied by civilians who kept the structures in repair, occasionally remodeling them to suit personal tastes. It is axiomatic that the only adobe buildings which remain intact are those with weatherproofed roofs.

The most well-intentioned, carefully planned, and systematically executed archeological excavation, however scientific, is likely to create difficult preservation problems. The principal cause of destruction is the result of archeological work which neither proceeds nor ends upon a single plane (fig. 4). This leaves the site partially filled at many different levels. Water collects in an area of two or three rooms and cuts its way to lower levels. Deeply excavated portions are subject to the pressure of damp fill. The loose, moisture absorbing fill in rooms transfers water to concealed walls and diminishes the strength of adobe mortar. In a site of any complexity, burrowing and tunneling are particularly destructive, even when backfilled, since the fills will not be as compact as the material which was removed.

Regardless of the condition in which standing walls are found, it must be remembered that the entire depth of the wall is seldom exposed, and that the exposed portions—which form a base over which the stabilization is being done—have been subject to nearly all the forces listed above, i.e., surface weathering, construction over loose fill or partial walls, breakdown or theft of wood members, erosion at the base, buckling from pressure, excavation, and backfill (fig. 5).

Frequently, cattle cause more damage to ruins in the Southwest than people. Until they were fenced and under the protection of the National Park Service, the well-preserved cliff dwellings in Walnut Canyon suffered greatly by trampling of cattle seeking shelter in the ruins. Herds of sheep and goats have created preservation problems at sites in Canyon de Chelly and Hubbell Trading Post. Ironically, large and imposing adobe wall sections of the corral at Fort Bowie were reduced to low mounds by open range livestock during a brief span of eight years—from the time of the Service's initial investigations to that of the enacting legislation which created the Historic Site. Erection of fences was given top priority immediately following acquisition.

FIGURE 4. Excavation neither proceeds nor ends on a single plane and often leaves walls unsupported (Kiva I, west perimeter, Chettro Kettle Ruin, Chaco Canyon National Monument, N. M.)

FIGURE 5. Clearing of depressed areas leaves them open to surface runoff and rapid destruction. (The Great Kiva, Kin Nahasbas Ruin, Chaco Canyon National Monument, N. M.)