Abstract

The fluting of Early Paleo-Indian bifaces, and failure rates therein, has reccived considerable archaeological attention – perhaps at a scale exceeding its relative importance. Nonetheless, if we are to properly evaluate explanations of the origins and function of fluting and the place of particular sites in Paleo-Indian settlement systems, estimates of failure rates are required. To date, such estimates have been largely used on experimental replication, and rarely on archaeological data, and it is suggested these estimates are inaccurate. In order to overcome these problems, three interrelated, yet independent, methods; of estimating fluting failure rates from archaeological data are developed and applied to assemblages from the Parkhill site in Ontario and the Windy City site in Maine. The consistency in the results obtained by these methods inspire faith in their relative accuracy. The results suggest failure rates, at least at these sites, were not on the scale often assumed by previous investigators, being on the order of only 10 to 12% percent or less.