They Mapped 150 Million Synapses, and Found No Sign of Stored Memories

A recent press account states this:

" Researchers have reconstructed a minuscule piece of the human brain down to the level of individual synapses, representing a giant leap forward for brain science. And we're not just talking about a few neurons here. This millimeter-sized cube contains a staggering 57,000 cells, 230 millimeters of tiny blood vessels, and nearly 150 million synaptic connections, all mapped out in glorious 3D detail."

We are told that the bit of brain tissue mapped was taken from an epilepsy patient, presumably in one of those operations in which brain tissue is removed to help prevent seizures. The brain tissue was divided into 5000 different sections, and then thoroughly photographed with an electron microscope. 

A web site allows access to some of the resulting photos and 3D models, and a scientific paper describing the data. What is cool is that upon looking at one of the photos, you can click with your mouse, and then drag around the image, which will rotate in 3D.

The page here allows you to see some neurons and their connections, and each neuron and its connections is shown in a separate color:

At the page here you can see a view of some dendrites and their little bumps called dendritic spines. Below is what you first see. By clicking on the right image, you can drag around in 3D the right part. 

Dendritic spines are the little bumps we see above

We can see from the scales at the bottom how very small are the structures shown. The dendritic spines shown above have a size of only about 1 micrometer, a millionth of a meter. The symbol for a micrometer is "µ."

Using the page here, you can see a synaptic gap between two connected synapses. The orange and green units on the left are synaptic boutons that correspond to the similar-looking orange and green units on the right. The synaptic clefts are about 300 to 700 nanometers wide. A nanometer is a billionth of a meter, or a thousandth of a micrometer. The symbol for a nanometer is nm. 

The investigation of this neural tissue is described in the paper here, entitled "A connectomic study of a petascale fragment of human cerebral cortex." It is interesting to see the discrepancy between the paper's lip-service to neuroscientist dogmas about a brain storage of memory, and the complete failure to supply any observations supporting such dogmas. 

"The most functionally significant aspect of cortical tissue is the synaptic connectivity that allows neurons to send and receive signals to and from other neurons. This wiring diagram is likely central to the way human brains store memory and give rise to behavior....Given the far greater variability in human experience, behavior, memory and genetics, and the fact that humans and other vertebrates have pools of identified neurons classes rather than individual identified neuron types, it will no doubt be challenging to compare neural circuits between brains....Even if the circuits differ in their particulars, it is possible that a metalogic for memory can be uncovered by looking at enough data, maybe in the future field of 'engramics.' To be sure, approaches to the profound questions of uncovering the meaning in neural circuit connectivity data are in their infancy, but it would seem to us that perhaps the best stimulus for making progress will be an abundance of actual data -- this petascale dataset is a start."

Those are all of the references the paper makes to memory, except for a reference to computer memory. The claim that studies looking for evidence of memory storage in brain tissue are in their infancy is incorrect, as is the claim that this study "is a start."  As I document in my post "They Stored and Studied Thousands of Brains, But Still Failed to Show Brains Store Memories," studies like this have been going on for decades.  And none of them have produced any evidence that memories are stored in brains. 

Could it be that the dendritic spines like the ones we see above are some kind of information representation system? There's no chance of that. For one thing, such dendritic spines lack any regularity such as you might have in an information representation system.  Dendritic spines appear with shapes and numbers as random as the buds coming from trees branches. Just as no one analyzing the buds coming from tree branches can detect any kind of code in which some symbolic information is being stored, no one analyzing the dendritic spines coming from dendrites can detect any kind of code in which symbolic information might be stored.  Just as the buds coming from tree branches have none of the regularity that we might find if symbolic information was being stored, the dendritic spines coming from dendrites have none of the regularity that we might find if symbolic information is being scored. And just as there is in nature no such thing as a tree branch bud reader, there is in the brain no such thing as a dendritic spine reader. And just as the buds coming out of tree branches don't last very long (lasting much less than a year), the dendritic spines coming out of dendrites don't last very long. For details see my post "Imaging of Dendritic Spines Hint That Brains Are Too Unstable to Store Memories for Decades."  The average lifetime of a dendritic spine is only about 120 days. 

Another example of the empty hand-waving of neuroscientists in regard to memory can be found in the paper here, entitled "Why not connectomics?" We have this example of conceptually empty hand-waving about memory storage:

"Brains can encode experiences and learned skills in a form that persists for decades or longer. The physical instantiation of such stable traces of activity is not known, but it seems likely to us that they are embodied in the same way intrinsic behaviors (such as reflexes) are: that is, in the specific pattern of connections between nerve cells. In this view, experience alters connections between nerve cells to record a memory for later recall. Both the sensory experience that lays down a memory and its later recall are indeed trains of action potentials, but in-between, and persisting for long periods, is a stable physical structural entity that holds that memory. In this sense, a map of all the things the brain has put to memory is found in the structure—the connectional map."

The first sentence is groundless dogma. There is no evidence that brains "can encode experiences and learned skills in a form that persists for decades or longer."  There is merely the fact that humans can have experiences and learn skills that they remember for decades.  The beginning of the second sentence is a confession that there is no understanding of how such a brain storage of memories can happen. The authors confess that "the physical instantiation of such stable traces of activity is not known,"  The claim that memories are stored by "the specific pattern of connections between nerve cells" is empty hand-waving, and the speculation stated is unbelievable. No one who has ever studied the connections between nerve cells (neurons) has ever seen anything like some symbolic pattern that could encode a record of human experiences or human learned skills or learned conceptual knowledge such as school learning.  The brain does not have any such thing as a connection pattern reader that could read and interpret such patterns if they existed.