This study examines the variability in Ocmulgee Fields ceramics from 2 nearly contemporaneous Upper Creek Indian sites at Horseshoe Bend National Military Park in central Alabama. The intent is to demonstrate that differences between the 2 pottery assemblages reflect differences in the social settings at the 2 sites. Frequencies of ceramic types and of selected attributes are tested for statistical significance using chi-square, and the strength of the assumed relationships are evaluated using 1 or more measures of association. The tests support a contention that ceramic vessels in use at the temporary, warrier-refugee settlement of Tohopeka were more frequently of utility forms and were generally larger and less decorated than vessels in use at the permanent, family-structured settlement of Nuyaka. It is suggested that similar studies of ceramics from sites for which the historical record is not as complete as it is for the Horseshoe Bend sites might reveal potential distinctions in social structure and behavioral patterning that would be developed primarily from the archaeological record.

Pomacea flagellata (Say), a large freshwater snail, was eaten by the ancient inhabitants of Tikal. Its shells appear in archaeological contexts datable to the entire circa 1500-year span of permanent occupation. Utilization was heaviest during the earliest and latest periods, a distributional pattern that seems directly related to population growth and decline. Comparison with the freshwater mussels and European land snails suggests that Pomacea snails were probably never more than a supplementary source of protein and calories. However, they could have been important in a marginal diet. Occasionally Pomacea were deposited in ceremonial contexts. They were most frequent during the later Early Classic period when they may have figured in a special votive complex.

Archaeology‘s relationship to anthropology in the United States has been one of a natural and beneficial alliance. Archaeologists are currently showing more of an interest in formal models drawn from outside anthropology, but the classification of American archaeology as a subdiscipline in anthropology generally remains unquestioned. We argue that at the present time archaeological research is being hindered by its institutionalized relationship to anthropology and its uncritical use of models from other disciplines. Archaeologists will make the greatest theoretical progress if they view their discipline as an autonomous technique with no a priori ties to sociocultural anthropology. Archaeology as a technique makes possible a truly interdisciplinary research base, but requires in turn a reorganization of research and training procedure as well as an academic restructuring.

The evidence for the production of tecali (travertine) vessels at Tollan, Hidalgo, Mexico is examined and some possible secondary uses for the resultant waste materials are considered. An attempt is also made to show how artifact reuse can confuse the archaeological record, possibly leading to false conclusions.

Our understanding of the archaeological record has been developed under the culture history paradigm. Its fundamental structure is shown to be stylistic; this characteristic, coupled with historical factors, is seen as the major reason why evolutionary processes have not been extensively employed in explaining cultural change. Consideration of an evolutionary approach suggests that such processes as natural selection have considerable explanatory potential, but it is also suggested that a substantial segment of the archaeological record is not best understood in terms of adaptation. The potential of an evolutionary approach cannot be realized without making a fundamental distinction between functions, accountable in terms of evolutionary processes, and style, accountable in terms of stochastic processes.

Squash remains from three Late Archaic archaeological sites constitute the earliest evidence for cultigens in eastern North America. The new data indicate the tropical cultigen, squash, was introduced into the area prior to the domestication of native plant resources.

The problem of discontinuity in the behavior of culture systems-the last stronghold of the anti-processualists-is discussed. Abrupt change in behavior can now be described in terms of smooth changes in the underlying causative factors by means of Rene Thom"s Theory of Elementary Catastrophes. The theory suggests insights not only into discontinuities with respect to time ("sudden" changes) but into the differentiation of forms as the result of bifurcations (morphogenesis). Although existing applications of the Theory in the social sciences lack quantitative precision, they offer a deeper understanding of crucial mechanisms of social evolution and, it is suggested, go far toward solving the discontinuity problem in archaeology.

A recent study of a combination open and closed sascab mine in Coba, Quintana Roo, Mexico revealed the techniques used to exploit its deposits of calcareous sand. Tool types used were determined by scars on the face of the deposit and the mine roof. Post production activities involving open sascaberas suggest their use as sunken gardens for both intensive horti-and arboriculture activities.

Geomorphologic and sediment-stratigraphic study of the Koster site has been carried out in the broader context of the lower Illinois Valley. Accumulation of reworked loess in an overdeepened tributary valley began at Koster shortly after 10,000 B.P., and continued through Holocene times with major sedimentary breaks. The Illinois floodplain began to stabilize ca. 5000 B.P. after rapid aggradation, but remained a dynamic environment that developed its present patterns after 2500 B.P. Valley-margin hillside vegetation was considerably more xeric during the periods 1200-950 B.P., 2100–1900 B.P., and ca. 9700–5000 B.P., with hillside woodland reduced to hill prairie or parkland ca. 8500–7700 B.P. These dramatic Holocene environmental changes suggest that interpretative archaeological models for cultural adaptations through time must consider the environment as a critical variable, rather than as a constant.

Meadowcroft Rockshelter is a deeply stratified multicomponent site in Washington County, southwestern Pennsylvania. The 11 well-defined stratigraphic units identified at the site span at least 16,000 years and perhaps 19,000 years of intermittent occupation by groups representing all of the major cultural stages/periods now recognized in northeastern North America. Throughout the extant sequence, the site served as a locus for hunting, collecting, and food-processing activities, which involved the seasonal exploitation of the immediately adjacent Cross Creek Valley and contiguous uplands. Presently, Meadowcroft Rockshelter represents one of the earliest well-dated evidences of man in the New World as well as the longest occupational sequence in the Western Hemisphere.

Little attention has been paid by archaeologists to the important problem of replicability of observational units. The studies presented in this paper involve examinations of discrepancy occurring at various levels in lithic and ceramic classification. Standardized typologies, as well as qualitative and quantitative attributes are considered. The results are discussed in terms of observer bias, influence of training, measurement error, and the implications for the statistical treatment of data.

The process of pattern recognition using data from historic sites is illustrated with data from the British colonial system. The Brunswick Pattern of Refuse Disposal monitors eighteenth century refuse disposal behavior patterns. The Carolina Pattern monitors artifact relationships from domestic occupation. The Frontier Pattern is seen on frontier sites as well as the area inside domestic ruins, and is characterized by a high architecture to kitchen artifact relationship. The formula concept of pattern recognition demonstrated by the Mean Ceramic Date Formula is a tool based on the recognition of highly regular patterns of variation in the popularity of ceramics through time. Such pattern recognition is foundational for historic site data to contribute to the explanation of culture processes. Historical archaeology has an as yet unrealized potential for contributing to method-refinement and theory building in archaeology generally. This is the exciting promise historical archaeology holds for the future.

The archaeology of the past two decades has become increasingly quantitative, computerized, statistical, and this is as it should be. All right-thinking archaeologists begin with samples and attempt to generalize about the populations from which their samples were drawn. Statistical theory has evolved to assist investigators in making just this important inferential step and archaeologists have increasingly turned to statistics to square their research with the canons of Science. But the statistical revolution in archaeology is not without its price. We must now face the fact that all applications of statistics to archaeology can no longer be applauded. The archaeological literature is badly polluted with misuses and outright abuses of statistical method and theory. This paper discusses some of these faulty applications and makes some recommendations which, if heeded, should improve the quality of quantitative methods in archaeology.

In 1958 and 1969 the Tokyo University expeditions excavated the Formative site of Las Haldas on the Peruvian coast as part of their multiyear investigation into the developmental processes of Andean civilization. The site, known for a temple complex consisting of 6 terraces and a circular sunken court, provides comparative data for the key highland Formative site of Kotosh. The results of the 1969 excavation refuted the earlier view that the temple complex was preceramic. The large quantity of organic refuse, the architectural build-up, and the radiocarbon dates suggest long, continuous occupation based on maritime economy. Las Haldas was a key component of a pan-Andean exchange system, which cross-cut major ecological zones, and must be considered in conjunction with agricultural populations in the nearby Casma Valley and farther inland.

Neutron activation analysis of a large number of samples of midwestern cherts shows systematic variation of trace-element proportions within several chert formations. This chemical variation parallels variation in physical properties of the cherts and can also be found between different outcrops of the same chert. Surface alterations of artifacts after they have been deposited in soils also appear to change the chemical properties of the materials to some degree. All of these factors will complicate attempts to characterize chert sources and assign artifacts to them on the basis of their trace-element compositions. Nevertheless, adequate sampling and analysis of cherts should allow discrimination of sources in most areas, since the patterning of variation between cherts appears to be meaningful.

The special problems of textile analysis in archaeology are reviewed. The present state of the art concerning identification of fibers, analysis of structure and technique, dyes, and analytical techniques is outlined. Suggestions are made for securing comparative collections and for recording archaeological specimens.