1. Barton, do I have to refer to you as “Key Master” from now on?
2. I decline the nomination, as the patent nerds might vote me off the island if they knew I had other alliances.
3. Yea, that is kinda what I was going after. I see what I think is the same conceptual confusion as well in the fact that all of the survey stuff in the generic-marks section involves survey evidence that is aimed at showing distinctiveness. The survey in the PRETZEL CRISPS case doesn’t seem to me to be at all relevant to whether PRETZEL CRISPS is a descriptive term or a generic term as the poll uses a category-brand binary rather than a genus-attribute binary. In contrast, in a case where the plain-meaning, non-TM sense of a word is clearly not descriptive (e.g., Pilades, as it refers to a person), then the survey data seem very relevant to providing genericide over time. Nobody is arguing that Pilades is a descriptive mark: its either inherently distinctive or not protectable.