An Estimate of How Much People Remember, Not of Underlying Cognitive Capacities

In my paper “How much do people remember? Some estimates of the quantity of Learned Information in long-term memory,” I tried to measure the amount of knowledge that people acquire and retain through experience. These were described as “black box” measurements of “functional information contents,” explicitly defined in the paper as “the amount of information transmitted from input to output with delay.” In his paper “Estimating human cognitive capacities: A response to Landauer,” (1988) Lawrence Hunter scathingly criticizes me for not providing upper bounds on underlying “component capacity” or “cognitive processing capacity.” In my paper, I had tried to make it clear that I did not intend to estimate such capacity bounds, offered some reasons why attempts to do so would be fruitless at this time, and discussed the interest and utility of the average information content estimates that I did provide. When an author is criticized for not addressing the same questions that the reader would have liked answered, it is hard to know what to do. It is tempting to merely apologize for the reader’s disappointment, or to urge more careful reading. But it would be unfortunate if my estimates of remembered information were construed as something they are not, and Hunter’s response proves that it is possible to misunderstand them rather dramatically. Let me try again. First I will briefly review what I did in the work reported in the paper, what the results mean and how they might be used. Then, I will comment, also briefly, on the alternative research goals proposed by Hunter. Finally, I’ll discuss a few technical points.