Pueblo pottery making: a study at the village of San Ildefonso

by Carl E. Guthe

As in any scientific problem, the first step must be to collect and classify the data; and although Pueblo sites are numbered in thousands, and an overwhelming amount of merely descriptive work must still be done, the outlines of a classification have been achieved. We know, for example, what sort of ruins are found in the San Juan drainage of northern New Mexico; what kinds of pottery occur along the Gila river in southern Arizona. But what relationship, genetically and in time, there existed between, say, Pueblo Bonito and Casa Grande, we do not know. The time element, or in other words the historical sequence of our material, remains in large part to be determined.

To set up an historical outline we must first determine the relative ages of the different ruins, and then estimate the size and distribution of the Pueblo tribes from the earliest times to the present. At the top, so to speak, of our series the problem is simple enough—we are acquainted with the present location of the tribes, and the various Spanish accounts tell us where they have been living during the past three hundred and fifty years. But for the prehistoric period (and everything in the Southwest prior to 1540 is prehistoric) we must rely almost wholly on such evidence as may be turned up by the archaeologist’s shovel, for of native written records there are none, nor can native legendary testimony be safely depended upon. This, of course, throws a heavy burden upon the archaeologist, a burden which is made heavier by the fact that stratigraphy of remains is so rarely found in the Pueblo country.

Stratigraphy, in other words the superposition of the more recent upon the more ancient, has been the Open Sesame to all the reconstructive sciences. The very framework of geology, for example, has been built up from stratigraphic observations. In archaeology, too, stratigraphy has revealed the sequence of the Stone Ages; made clear the development of the early Mediterranean cultures, and the rise of predynastic civilization in Egypt. Therefore, as has just been said, the general lack of stratigraphic conditions in the Southwest renders the task of the student a particularly hard one. For some reason, not yet clearly understood, the Pueblos ancient and modern were very prone to shift from one dwelling place to another, and a site once abandoned was seldom reoccupied. Although their houses were of the most permanent construction, and their agricultural life should have tended to render them solidly sedentary, they moved about to a surprising extent. The result of this is that one seldom finds a ruin which was lived in for more than a few decades or, at most, centuries; and few have so far been discovered to contain superimposed remains illustrating any long period of development. Where such evidence is so rare, what can be found naturally becomes of the greatest importance; hence the recent diligent search for, and excavation of, such sites as show signs of long occupancy.

The choice of Pecos for investigation by Phillips Academy was due to the above considerations. The ruin was a large one, was occupied at the time of the Discovery and was not abandoned until 1838. A surface examination also showed that it must have been tenanted for a long time prior to the Conquest because its mounds were scattered over with potsherds not only of recent date, but also of several distinct prehistoric types, each one well enough known to students, but whose relative ages were entirely a matter of conjecture. It was hoped, therefore, that excavation might disclose some definite cases of superposition, and that several prehistoric periods might thereby be arranged in their proper chronological order.[1]

The results have been more than satisfactory. Pecos proved to have been built on the edge of a sharp-sided mesa, a fact not suspected before digging began, because the rubbish from the town had heaped up to such an extent against the original cliff as completely to mask its steepness. The first inhabitants naturally threw their refuse over the edge of the mesa, their descendants added to the accumulation, and the process continued down the centuries until there grew up a midden of enormous extent and, for the Southwest, of unusually great depth. It is stratified as neatly as a layer-cake.

When the exploratory trenches revealed the size and probable importance of the Pecos rubbish heap, all other projects were postponed, and two full field-seasons were devoted to the meticulous dissection of large areas of the deepest deposit. At frequent intervals stratigraphic tests were made, in which all the specimens from each successive stratum were kept separate and shipped to the Museum for study. It was found that many changes in culture had taken place during the long occupancy of Pecos; in the stone and bone implements, in the pipes, and in burial customs. But the most abundant, the most easily gathered, and the most readily interpreted evidence of cultural change was offered by the thousands of pottery fragments that filled the mound from subsoil to surface.

We have been able to recognize about twenty distinct wares, to arrange them into eight chronological groups, and to determine the exact sequence of these groups. This information, derived from the stratigraphic study of the pottery in the mounds, has been of the greatest value. Its application has been both local and non-local. In the excavations that we have since carried on at Pecos it has enabled us to date relatively to each other the various kivas, cemeteries, and small refuse mounds that occurred on the mesa top, and also to unravel much more confidently than we would otherwise have been able to do, the extraordinarily complex jumble of ruined, torn-down, stone-robbed, rebuilt, abandoned, and reoccupied rooms that we encountered when we attacked the pueblo itself. As helpful as has been the knowledge of the sequence of the pottery types in working out the details of local archaeology, its usefulness in that regard is small as compared with the flood of light which has been thrown on much larger and more vital problems. It has just been stated that some twenty types of pottery were identified; the majority of these are not peculiar to Pecos; many of them occur throughout large areas in the Rio Grande drainage; and so we are now able, in most cases by a hasty examination of the surface sherds, to assign to its proper place in the chronological series any ruin at which our types are present. Thus mere reconnaissance (a cheap and rapid undertaking) now serves to make clear the major outlines of Rio Grande archaeology. But the usefulness of the stratigraphic studies at Pecos does not end with the territory in which the Pecos types of pottery are found, for Pecos, because of its size and because of its situation on the main route between the Pueblo country to the West and the buffalo ranges to the East, was an important trade centre.

From its mounds we have taken potsherds from almost all parts of the Southwest, shells from the Pacific, spindle-whorls from Central Mexico, as well as pottery and stone objects from the Mississippi drainage. The importance of such finds is evident; every sherd from an outside culture found in a datable stratum at Pecos helps to fit into our general chronological scheme the culture from which it came; as conversely, does every Pecos or even Rio Grande sherd that turns up beyond the limits of the Rio Grande. Only a start has been made, but enough has already come to light at Pecos and at such stratified sites as have been excavated by other institutions, to provide us with a surprisingly full knowledge of the rise and growth of the Pueblo civilization. Ten years ago it would have been hard to believe how much could be accomplished by the stratigraphic work carried on by Nelson in the Galisteo Basin, Hodge at Hawikuh, Judd at Pueblo Bonito, Morris at Aztec, Guernsey in the Kayenta country, and the writer and his associates at Pecos.

The work, as I have said, has just begun, but the prospects are bright. Success will depend, as in any such endeavor, upon intelligent excavation, careful collection of data, and accurate observation of specimens, but the investigator cannot hope to derive the best results from his labors if he does not hold to a very broad view of his field. He must familiarize himself not only with the material of the locality he is working in but must also know as much as possible about that of other regions, for it is of the last importance that trade-objects be recognized wherever they occur. Of all trade-objects, vessels and potsherds are likely to be the commonest and most easily recognizable.

When the members of the Pecos expedition realized how vitally important was to be a close knowledge of the pottery, not only of that particular site, but also of the Rio Grande in general, and, indeed, of the entire Southwest, they devoted a large part of their time, both in the field and at the Museum, to the study of ceramics. A difficulty was at once encountered in our ignorance of the technique of Pueblo pottery making. We were constantly obliged to compare the paints and slips of different wares, to decide whether observed variations were technically fundamental, or were merely accidents of clay-mixing or firing. No full published accounts existed, nor had any of us had the opportunity for more than casual observation of potters at work. Dr. Guthe, accordingly, spent the month of August, 1921, at San Ildefonso in making detailed studies. His results are published in the hope that they may be of use to Southwestern archaeologists and ethnologists, and also to students of the primitive pottery of other areas who may be in need of material for comparative purposes.