It’s quite possible that Alzheimer’s patients of the future will be equipped with memory prosthetics derived from the devices being invented in Berger’s lab today. His work began with delicate electrodes inserted into a rat’s hippocampus, the brain structure responsible for encoding memory. Berger first deciphered the relationship between the input signals from neurons that process a brief learning experience—for example, which lever a rat should press to gain a sip of sugar water—and the output signals from neurons that send the information on to be stored as a memory.
Once he had mapped the correlations between the two electrical patterns, Berger could record an input signal and predict the output signal—in other words, the memory. He didn’t need to know which part of the input pattern coded for the dimensions of the lever or for the taste of the sweet reward. He simply mathematically generated the output signal and sent it to the memory-storage neurons. “It’s like translating Russian to Chinese when you don’t know either language,” Berger says. “We don’t want to know either language; we just want to know how this pattern becomes that pattern.”
Berger proved that he could implant the memory of the lever-and-reward test in a rat with a damaged hippocampus that was unable to form memories on its own. Even more remarkable, he implanted the memory in a rat that had never before undergone the test or seen the levers. The rat entered the test chamber for the first time, pressed the correct lever, and sucked down the sweet nectar.