Protocols [lab experiments] certainly tell us something, but their web has not been precise enough to catch the important ingredients that go into cognition, memory, perception and attention, to say nothing of intention, desire, self-esteem and all the other variables of the rhetorical situation.
