It has recently been argued by Gess (2001) that the long vowels resulting from the compensatory lengthening that emerged in the wake of preconsonantal /s/-deletion in Old French had all been shortened by the sixteenth century. Given that many of these long vowels are still present in Canadian French, this conclusion cannot possibly be correct. What will be shown here is precisely how Gess' methodology led him to obtain such counterfactual results.