A singular consequence of the materialist-mechanical metaphysics that permeates our culture and our sciences is that we commonly hold basic beliefs that are abject nonsense. One such belief is the almost ubiquitous one — among ordinary folks as well as neuroscientists and surprisingly many philosophers — that the brain "stores" memories. The fact is that the brain doesn’t store memories, and can’t store memories.
It has been known for the better part of a century that certain structures in the brain are associated with memory. The amygdala and the hippocampus in the temporal lobe, and some adjacent cortical regions, have been shown to be associated with the act of remembering in animals and humans. The research is fascinating and important, and in my own work as a neurosurgeon I have to be aware of these regions (especially the hippocampus and the fornix and mammillary bodies, to which the hippocampus projects). During surgery, injury to these critical structures (if bilateral) can leave a patient incapable of forming new memories, which is a crippling disability.
But these physiological facts do not imply that the brain stores memories in the hippocampus or amygdala or elsewhere. How so?
It’s helpful to begin by considering what memory is — memory is retained knowledge. Knowledge is the set of true propositions. Note that neither memory nor knowledge nor propositions are inherently physical. They are psychological entities, not physical things. Certainly memories aren’t little packets of protein or lipid stuffed into a handy gyrus, ready for retrieval when needed for the math quiz.
The brain is a physical thing. A memory is a psychological thing. A psychological thing obviously can’t be "stored" in the same way a physical thing can. It’s not clear how the term "store" could even apply to a psychological thing.
Now you may believe — as most neuroscientists and too many philosophers (who should know better) mistakenly believe — that although of course memories aren’t "stored" in brain tissue per se, engrams of memories are stored in the brain, and are retrieved when we remember the knowledge encoded in the engram. Indeed neuroscientists believe that they have found things in the brain very much like engrams of some sort, that encode a memory like a code encodes a message.
But that too is nonsense. To see why, consider a hypothetical "engram" of your grandmother’s lovely face that "codes" for your memory of her appearance. Imagine that the memory engram is safely tucked into a corner of your superior temporal gyrus, and you desire to remember Nana’s face. As noted above, your memory itself obviously is not in the gyrus or in the engram. It doesn’t even make any sense to say a memory is stored in a lump of brain. But, you say, that’s just a silly little misunderstanding. What you really mean to say is that the memory is encoded there, and it must be accessed and retrieved, and it is in that sense that the memory is stored. It is the engram, you say, not the memory itself, that is stored.
But there is a real problem with that view. As you try to remember Nana’s face, you must then locate the engram of the memory, which of course requires that you (unconsciously) must remember where in your brain Nana’s face engram is stored — was it the superior temporal gyrus or the middle temporal gyrus? Was it the left temporal lobe or the right temporal lobe? So this retrieval of the Nana memory via the engram requires another memory (call it the "Nana engram location memory"), which must itself be encoded somewhere in your brain. To access the memory for the location of the engram of Nana, you must access a memory for the engram for the location for the engram of Nana. And obviously you must first remember the location of the Nana engram location memory, which presupposes another engram whose location must be remembered. Ad infinitum.
Now imagine that by some miracle (materialist metaphysics always demands miracles) you are able to surmount infinite regress and locate the engram for Nana’s face in your superior temporal gyrus (like finding your keys by serendipity!). Whew! But don’t deceive yourself — this doesn’t solve your problem in the least. Because now you have to decode the engram itself. The engram would undoubtedly take the form of brain tissue — a particular array of proteins, or dendrites or axons, or an electrochemical gradient of some specific sort — that would mean "memory of Nana’s face." But how can an electrochemical gradient represent a face? Certainly an electrochemical gradient doesn’t look like grandma — and even if it did, you’d have to have a little tiny eye in your brain to see it to recognize that it looked like grandma. Whatever form the engram takes must be a code, and you must then have a key to the code, stored in your brain just like the Nana memory is stored. But then you must remember where the key to the code is stored, which is itself another memory which must be stored and remembered. And to remember the location of a location for the key for the code for the engram requires another engram to remember the location of the location code, which must be located and decoded, which requires another key engram which you now must locate…
And if you think that remembering your grandmother’s face via an engram in your brain entails infinite regress, consider the conundrum of remembering a concept, rather than a face. How, pray tell, can the concept of your grandma’s justice or her mercy or her cynicism be encoded in an engram? The quality of mercy is not strained, nor can it be encoded. How many dendrites and axons for mercy?
You see the nonsense.
To assert that memories are stored in the brain is gibberish. And don’t fall for the materialist invocation of promissory materialism — "It’s just a limitation of our current scientific knowledge, and we promise that science will solve the problem in due time." The assertion that the brain stores memories is logical nonsense that doesn’t even rise to the level of empirical testability.
How then, you reasonably ask, can we explain the obvious dependence of memory on brain structure and function? While it is obvious that the memories aren’t stored, it does seem that some parts of the brain are necessary ordinarily for memory. And that’s certainly true. But necessary does not mean sufficient. There is a rough correspondence between activity in certain regions of the brain and the exercise of certain mental powers. That is what cognitive neuroscientists properly study. In some cases the correspondence between brain and memory is one of tight necessity — the brain must have a specific activity for memory to be exercised. But the brain activity is not the same thing as the memory nor does it make any sense at all to say the brain activity codes for the memory or that the brain stores the memory.
What this all implies is that only some kind of dualism can provide a coherent understanding of the mind. But dualism is a many-headed hydra, and I don’t think that Cartesian dualism or property dualism or epiphenomenalism or computational theories of the mind (which are inherently dualistic) explain things well either.
I hew to Thomistic dualism, which is a coherent view of the mind that takes an Aristotelian perspective and for which the participation of the brain in memory is not problematic at all.