Reconstruction
Meadowcroft Rockshelter is not the remains of one meal or settlement but a layered archive of repeated visits to the Cross Creek valley in southwestern Pennsylvania. Excavations begun in 1973 exposed eleven major stratigraphic units containing hearths, fire floors, pits, stone tools, basketry, cordage, and enormous collections of plant and animal material. The shelter was apparently used intermittently rather than as a permanent home: mobile groups stopped there to maintain tools, gather resources, process food, and camp beneath the sandstone overhang.
The deepest artifact-bearing deposits, especially Stratum IIa, made Meadowcroft central to the debate over settlement of the Americas before Clovis. Finds include small blades, bifaces, retouched flakes, and the unfluted Miller Lanceolate point. More than fifty radiocarbon determinations support a long sequence, but the oldest measurements have remained controversial. Critics proposed that ancient carbon from local coal or displaced charcoal made samples appear too old, and they noted that the sparse identified plants and animals in the lowest deposits resemble temperate woodland species. Micromorphological study found no evidence that groundwater contaminated the early sediments, strengthening the excavators’ case without eliminating every chronological or associational question.
No precise menu can be reconstructed for Meadowcroft’s earliest visitors. Lower deposits produced only a small number of species-level faunal identifications, including white-tailed deer, southern flying squirrel, eastern chipmunk, and passenger pigeon. Their presence does not by itself prove consumption because animals, predators, and natural processes can introduce remains into a shelter. Hearths, a roasting pit, cutting implements, and refuse or storage features provide direct evidence of food-related activity, but assigning particular species to particular meals remains an inference.
Evidence becomes richer in later Archaic deposits, after the period emphasized by this card. White-tailed deer, elk, turkey, smaller animals, and freshwater mussels occur alongside probable plant foods such as acorns, hickory nuts, walnuts, hackberries, blackberries or raspberries, cherries, grapes, and goosefoot. Associations among some of these remains, hearths, pits, and plant residues in plaited containers indicate gathering, heating, processing, and storage. These later finds should not be projected unchanged into the disputed pre-Clovis occupation, but they reveal how the same sheltered place became part of a broad-spectrum foraging landscape over millennia.
Meadowcroft’s importance therefore lies less in a single ancient dish than in its record of flexible food procurement. It challenges older portraits of early eastern North Americans as exclusively big-game hunters and demonstrates the value of fine excavation, flotation, faunal analysis, and radiocarbon dating. Wild nuts, berries, game, and river foods remain familiar regional resources today, although direct culinary continuity from Meadowcroft’s earliest occupants cannot be demonstrated.
Historical context
The upper Ohio Valley was passing from the final millennia of the Ice Age into the warmer Holocene. The Laurentide Ice Sheet had retreated north of southwestern Pennsylvania, but glacial meltwater, shifting forests, and rapid climatic changes continued to reshape the Northeast. The Younger Dryas cold interval occurred approximately 10,900–9700 BCE. Human populations were probably sparse and highly mobile, using rivers and tributary valleys as travel corridors. Clovis fluted-point technology appeared across much of North America around 11,100–10,800 BCE, while sites such as Meadowcroft contributed to the still-debated case for earlier settlement.
Evidence
Archaeological
StrongCarefully recorded stratigraphy directly preserves hearths, fire floors, pits, lithic artifacts, perishable artifacts, and occupation surfaces from repeated visits to the shelter.
Archaeobotanical
LimitedPlant remains are abundant across the complete sequence, and later Archaic food plants are comparatively well represented. Identified remains from the oldest deposits are sparse, however, and their association with human consumption is not always secure.
Zooarchaeological
LimitedThe site produced a very large faunal assemblage, but few specimens from the lowest deposits were identified to species. Natural accumulation and predator activity complicate interpretation of those remains as food refuse.
Food identification
LimitedFood-related activity is supported, but no specific menu can be directly reconstructed for the earliest occupation. Better-supported lists of nuts, fruits, game, and mussels principally derive from later deposits.
Dating
ModerateMore than fifty radiocarbon dates form an internally ordered chronology, but the oldest dates remain disputed because of possible old-carbon contamination, uncertain charcoal associations, and paleoenvironmental inconsistencies.
Preparation method
ProbableHearths, fire floors, a roasting pit, tools, and storage or refuse features support heating and processing food. Particular cooking methods and recipes are reasonable reconstructions rather than direct observations.
Geographic attribution
StrongThe rockshelter and excavated deposits are securely located above Cross Creek in Washington County, southwestern Pennsylvania.
Historical interpretation
InterpretiveIntermittent use by mobile foragers is supported by the site assemblage and setting, but seasonality, group size, resource priorities, and the relationship of the earliest component to other pre-Clovis traditions remain interpretive.
Sources
- 1.J. M. Adovasio, J. D. Gunn, J. Donahue, and R. Stuckenrath (1978). Meadowcroft Rockshelter, 1977: An Overview. American Antiquity 43(4): 632–651. doi:https://doi.org/10.2307/279496Primary archaeology
- 2.J. M. Adovasio, J. Donahue, and R. Stuckenrath (1990). The Meadowcroft Rockshelter Radiocarbon Chronology 1975–1990. American Antiquity 55(2): 348–354. www.cambridge.org/core/journals/american-antiquiScientific literature
- 3.Paul Goldberg and Trina Arpin (1999). Micromorphological Analysis of Sediments from Meadowcroft Rockshelter, Pennsylvania: Implications for Radiocarbon Dating. Journal of Field Archaeology 26(3): 325–342. doi:https://doi.org/10.1179/jfa.1999.26.3.325Scientific literature
- 4.Kenneth B. Tankersley and Cheryl Ann Munson (1992). Comments on the Meadowcroft Rockshelter Radiocarbon Chronology and the Recognition of Coal Contaminants. American Antiquity 57(2): 321–326. doi:https://doi.org/10.2307/280736Scientific literature
- 5.Jonathan C. Lothrop, Darrin L. Lowery, Arthur E. Spiess, and Christopher J. Ellis (2016). Early Human Settlement of Northeastern North America. PaleoAmerica 2(3): 192–251. doi:https://doi.org/10.1080/20555563.2016.1212178Modern synthesis
- 6.Nathaniel R. Kitchel and Madeline E. Mackie (2022). Plants and Subsistence during the Fluted-Point Period of the Northeast. American Antiquity 87(2): 368–376. doi:https://doi.org/10.1017/aaq.2021.125Scientific literature
Limitations
- moderatedate start
An initial occupation near 16,000 BCE is possible within published calibrated ranges, but Meadowcroft’s earliest dates remain disputed. Some syntheses accept only a more conservative human presence around 14,500 years before present, while the excavators argue for occupation approaching 19,000 calendar years ago.
- minordate end
The date 10,000 BCE may delimit this card’s selected late-Pleistocene episode, but it is not the end of Meadowcroft’s archaeological sequence. Intermittent use continued through the Holocene and into recent centuries.
