Recently, researcher John Hart, also of the New York State Museum, looked at the Roundtop collections. He was interested for several reasons, but primarily because archaeological techniques have improved in the 35 years since the first excavations. Foremost in Hart's mind was the fact that today, accelerator mass spectrometry (or AMS) radiocarbon dating now allows scientists to radiocarbon date much tinier pieces of carbon; and such dating had been used successfully on other occurrences of corns, beans and squash in the Americas, Hart knew that corn, at least, had been found in earlier contexts in the northeast, but he wanted to confirm the combination of corn, beans, and squash as being present at such an early date.
Much to Hart's astonishment, the carbon dates on the beans, corn, and squash came back around AD 1300, two hundred years and more later than the original study showed. Because his curiosity was piqued, he and research associate Margaret Scarry submitted beans from a handful of other sites in the American northeast, and none of them were dated much before AD 1300. Reinvestigation of the ceramics in Features 30 and 35 revealed that while the ceramics in Feature 30 were of the Early Owasco type; those in Feature 35 belonged to Late Owasco, and matched the AD 1300 date.
Bottom line? Beans do not show up in archaeological sites in the American northeast until AD 1300, and corn-beans-squash intercropping was not part of Native American culture in the American northeast until that time.
This is the way science works, or how it should work. We continue to work on the technology, and sharpen our understanding of the past as the years go on. But this kind of analysis and reanalysis and improved understanding can only happen if we are careful excavators, if we are willing to share our data openly, and if we can stash our egos where they belong: in the bottom of a storage pit.

