In the 1950's, Karl Lashley dedicated his research to the search for the center of learning in the mammalian cortex. He called this proposed learning center the 'engram', and believed that it was a specialized nucleus somewhere in the prefrontal lobes. When, after exhaustive experimentation involving lesioning the cortex and searching for learning defecits, he failed to uncover such a brain structure, he modified his theory to propose that learning takes place diffusely throughout the cortex, an idea called the Theory of Mass Action.

Many of the things that are difficult to account for presently in terms of physiology may work in this way--things like conscience, self-awareness, and will cannot be disproved to be functions of the nervous system just because there is no single identifiable brain structure responsible for their occurrence. Their presence may rather occur in neural circuits scattered throughout the brain, some interconnected with other "I-function" circuits, others with regions of the brain relating to emotion and other sensations. No-one can stick an electrode into any one area of the brain and stimulate feelings of consciousness, nor lesion a particular gyrus and turn off free will.

In the years following Lashley's work, many functions of memory and learning have been localized to the hippocampus and the process of long term potentiation. Maybe eventually all of the things that we are worried about defining as neural functions will be localized too...but it won't bother me too much if they aren't.

Nice that someone else has read Lashley. And nice suggestion as to what the difficult problems are and why. Localization is real, in the sense that different local regions of the brain do different things ... but what is localized in any given region has rarely (never?) turned out to be what people were looking to localize. Is a bit like roses and tulips discussed in class today: yes, they are different; yes, the brain distinguishes them, but it does so by patterns of activity distributed across lots of things. Maybe we shouldn't try and localize "the things we are worried about", but rather see what local regions DO and expect "the things we are worried about" to emerge from the interactions of regions doing THOSE things? PG