Quality Controls: The Role of Self-Corrective Science in Explorations of Primate Memory Systems.

In 1978, Mort Mishkin published a landmark paper describing a monkey model of H.M.'s dense, global amnesia. It depended on a combined removal of the amygdala and hippocampus (the A + H lesion) and a memory test called delayed nonmatching-to-sample (DNMS). My first project examined whether the impairment Mishkin had found in visual memory generalized to tactual stimuli. However, to gain access to the hippocampus and amygdala with 1980s surgical methods, we had to remove the underlying cortex. When we were able to test the effects of bilateral removal of that underlying cortex (the entorhinal and perirhinal cortex, or "rhinal cortex" for short) we obtained a dramatic result. This so-called "control" lesion caused a profound impairment on the DNMS task. A few years later, excitotoxic A + H lesions, which left the rhinal cortex intact, confirmed that removal of the cortical "impediments" had caused the entire memory impairment that Mishkin had observed. These results: (1) forced a reconsideration of the monkey model of global anterograde amnesia; (2) spurred study of the independent contributions of the amygdala, hippocampus, and perirhinal cortex to cognition; and (3) led to the realization that the DNMS task did not test the kinds of memory that H.M. lost after his surgery.