Microscopes Will Never Find the Slightest Trace of Learned Information in a Human Brain

A recent story at the LiveScience.com site has the title "Could we ever retrieve memories from a dead person's brain?" The story has many  inaccurate claims, and fails to tell us the most important facts that are relevant to the question being considered. The subtitle of the article makes this claim: "Neuroscientists have identified the physical locations where memories are stored in the brain."  No, they have not done any such thing, and the article fails to present any good evidence that any such thing was done. 

We have this claim about a method to retrieve a memory from a brain:

"With today's technology, retrieving memories might go something like this. First, identify the set of brain cells, or neurons, that encoded a specific memory in the brain and understand how they are connected. Then, activate those neurons to create an approximate neural network, a machine learning algorithm that mimics the way the brain works."

This does not make any sense as an idea about how one would go about trying to start to read a memory from a brain. The first step in such a process would be to use microscopes to look for any speck of a trace of learned information in a brain. No one has ever succeeded in doing any such thing. Microscopic examination of brain tissue has never revealed a single word anyone ever learned. Microscopic examination of brain tissue has never revealed a single letter or character of anything anyone previously learned, and has never revealed a single pixel (an image dot) of anything anyone ever previously saw. Neural networks are a misnamed type of computer technology that do not actually mimic the brain and its physical shortfalls. 

We have quotes by a neuroscientist (Don Arnold) doing some vague hand-waving and speaking as if he knew things he does not actually know. We read,  "Memories are encoded by groups of neurons, Arnold said." That's the kind of vacuous, vague hand-waving that someone may use when he lacks any actual knowledge of how a brain could store memories.  When people make big important-sounding claims that are not well-supported by evidence, they typically speak in a kind of vague, hand-waving way.  For example, when asked where there existed the weapons of mass destruction that the US claimed were in Iraq before invading it in 2003, US defense secretary Donald Rumsfeld said this:

"We know where they are. They're in the area around Tikrit and Baghdad and east, west, south and north somewhat."

Such weapons of mass destruction were never found in Iraq in 2003 or in the next twenty years. 

Reminding me of the Rumsfeld statement, the LiveScience article claims that long-term memories are formed in the hippocampus, a claim not backed up any robust evidence. The main research paper on the hippocampus and memory is the paper "Memory Outcome after Selective Amygdalohippocampectomy: A Study in 140 Patients with Temporal Lobe Epilepsy." That paper gives memory scores for 140 patients who almost all had the hippocampus removed to stop seizures.  Using the term "en bloc" which means "in its entirety" and the term "resected" which means "cut out," the paper states, "The hippocampus and the parahippocampal gyrus were usually resected en bloc."  The "Memory Outcome after Selective Amygdalohippocampectomy" paper does not use the word "amnesia" to describe the results. That paper gives memory scores that merely show only a modest decline in memory performance.  The paper states, "Nonverbal memory performance is slightly impaired preoperatively in both groups, with no apparent worsening attributable to surgery."  In fact, Table 3 of the paper informs us that a lack of any significant change in memory performance after removal of the hippocampus was far more common than a decline in memory performance, and that a substantial number of the patients improved their memory performance after their hippocampus was removed. 

The LiveScience article tells us, "Other parts of the brain store different aspects of a memory, like emotions and other sensory details, according to the Cleveland Clinic." We are referred to a page on some web site of the Cleveland Clinic that has no named author, and is the kind of article that you might get if you asked ChatGPT about how memory works. That Cleveland Clinic page does not actually make the claim that the LiveScience article attributes to it. The Cleveland Clinic does not claim that different aspects of a memory are stored in different places, but merely claims that other parts of a brain "participate in memory processes."

The LiveScience article then makes a false claim, repeating a groundless achievement legend. It states this:

"Neuroscientists have identified engrams in the hippocampuses of mouse brains. For instance, in a 2012 study published in the journal Nature, researchers found the specific brain cells associated with a memory of an experience that induced fear." 

No, there does any exist any robust evidence for engrams (neural storage places of memories) in any animal.  The paper the LiveScience article links to is the 2012 paper “Optogenetic stimulation of a hippocampal engram activates fear memory recall.” That is a very low-quality paper guilty of several bad examples of Questionable Research Practices. We see in Figure 3 of that paper that inadequate sample sizes were used. The number of animals listed in that figure (during different parts of the experiments) are 12, 12, 12, 5, and 6, for an average of 9.4. That is not anything like what would be needed for a moderately convincing result, which would be a minimum of 15 or 20 animals for each study group, and probably more. The experiment relied crucially on judgments of fear produced by manual assessments of freezing behavior, which were not corroborated by any other technique such as heart-rate measurement. All mouse research papers relying on "freezing behavior" judgments are junk-science papers, for reasons I discuss in my post here, "All Papers Relying on Rodent 'Freezing Behavior' Estimations Are Junk Science."  The 2012 study does not describe in detail any effective blinding protocol, which is another bad defect.  The study involved stimulating certain cells in the brains of mice, with something called optogenetic stimulation. The authors have assumed that when mice freeze after stimulation, that this is a sign that they are recalling some fear memory stored in the part of the brain being stimulated. What the authors neglect to tell us is that stimulation of quite a few regions of a rodent brain will produce freezing behavior. So there is actually no reason for assuming that a fear memory was being recalled when the stimulation occurs. 

There does not exist any observational or experimental support for the existence of engrams (memory storage places) in any animal. Some papers have claimed to have produced such evidence, but their claims do not stand up to critical scrutiny. Papers claiming to produce such evidence are generally guilty of multiple types of Questionable Research Practices such as way-too-small study group sizes, lack of a blinding protocol, and the use of one or more unreliable techniques for judging memory performance. 

The LiveScience article then gives us this bit of excuse-making for why no one has ever read a memory from a brain: "The retrieval of a dead person’s memories is further complicated because the discrete parts of a memory are dispersed throughout the brain; for instance sensory details that can also be stored in the parietal lobe and sensory cortex." This is an appeal to the theory that a single memory is stored not in one tiny part of the brain but in multiple scattered parts of the brain. There is no evidence for such a theory, and the theory makes things worse for the person claiming that the brain stores memories, for reasons I discuss in my post "Why the 'A Memory Is Stored Throughout the Brain' Idea Makes Things Much Worse."  If a single memory were to be stored in multiple locations in the brain, then finding all of those locations and assembling them instantly (in a brain without any addresses or indexes or sorting) would be something even more impossible to explain than imagining that the memory existed in a single spot that was instantly found. 

Giving us its second example of claiming a cited source said something that it did not actually say, the LiveScience article states, "Neurons within a given engram are connected through synapses, the spaces between neurons where electrochemical signals travel, according to the National Library of Medicine."  The page that it links to does not ever use the word "engram," and does not refer to either memory or learning. The LiveScience article claims that according to the neuroscientist Arnold, "there is evidence that memories move to different locations as they are consolidated in the brain." There is no robust evidence of any such thing. Neuroscientists have no credible evidence of memories being stored in any part of the brain of any organism, and they do not have any decent evidence of a memory moving around from one part of the brain to another. 

Arnold is quoted as saying, "You get this sort of cascade of neurons that encode these different things, and each one of them is connected in this engram." That is hand-waving. Scientists lack any robust evidence of any such thing as an engram or an encoding of learned information or experiences in the brain. No scientist has a credible detailed theory of how such encoding could occur.  An ocean of difficulties arises when you start to consider the endless problems that would arise when trying to translate human learned knowledge and experiences into brain states through any imaginable system of encoding. Part of the problem is the extreme variety of things that people can learn and experience (concepts, facts, theories, visual  experiences, auditory experiences, smell experiences, taste experiences, pain experiences, touch experiences, and emotional reactions), meaning there could be no simple encoding scheme (something as simple as the genetic code) that could handle even a tenth of all the types of memories people can form. 

We have no discussion of some of the chief facts relevant to the topic discussed. Some of these facts are below:

(1) Human brain tissue has already been exhaustively studied at very high microscopic resolutions. My post "They Stored and Studied Thousands of Brains, But Still Failed to Show Brains Store Memories" discusses how places such as the Lieber Institute have microscopically studied thousands of brains, most of which were preserved very soon after death. The same post describes how Denmark's University of Odense has stored more than 9000 brains, microscopically examining a large fraction of them. Very much healthy brain tissue just-extracted from living patients has been microscopically examined, because normal brain tissue is often extracted from epilepsy patients when operations are done to prevent intractable seizures resistant to medicine. 

(2) Despite all of that microscopic examination, no one has ever found the slightest trace of any learned information by microscopically examining a brain. Microscopic examination of brain tissue has never revealed a single letter or character of anything anyone learned, and has never revealed a single pixel of anything anyone ever saw. It isn't just that no one ever found anything like "The US has 50 states" by microscopically examining brain tissue; it's that no ever found a U or an S from microscopically examining brain tissue. 

(3) The discovery of a bit of learned information from microscopically examining brain tissue is something that would be many times easier to do than the "recreation of a full memory" imagined by the LiveScience article, but neither of these things has occurred. 

(4) Modern microscopic techniques are powerful enough to discover traces of learned information in the brain if they existed, but no such discovery has occurred. 

(5) Nothing in the brain looks like any type of mechanism for storing learned information or storing memories.  Nothing in the brain looks like any type of apparatus for writing learned  information. 

(6) Nothing in the brain looks like any type of mechanism for reading a stored memory. We can imagine how some organism's brain might physically look like something capable of reading information from a particular spot, by means of something like a moving cursor or moving reading component. The brain has no such thing, and the brain has no moving anatomical parts. 

(7) The places claimed to be sites of brain memory storage (synapses) are unstable places of high molecular turnover, where all the proteins last for only a few weeks or less. There is no credible theory of how places so unstable could be storage places for memories that can reliably last for 60 years. 

(8) There is nothing in the brain that looks like learned information stored according to some systematic format that humans understand or do not understand. Even when scientists cannot figure out a code used to store information, they often can detect hallmarks of encoded information. For example, long before Europeans were able to decipher how hieroglyphics worked, they were able to see a repetition of symbolic tokens that persuaded them that some type of coding system was being used. Nothing like that can be seen in the brain. We see zero signs that synapses or dendritic spines are any such things as encoded information. 

(9) Many humans can remember with perfect accuracy very long bodies of text, but synapses in the brain do not reliably transmit information. An individual chemical synapse transmits an action potential with a reliability of only 50% or less, as little as 10%. A recall of long bodies of text would require a traversal of very many chemical synapses. A scientific paper says, "In the cortex, individual synapses seem to be extremely unreliable: the probability of transmitter release in response to a single action potential can be as low as 0.1 or lower."

Below is a diagram from the paper "Materials Advances Through Aberration-Corrected Electron Microscopy." We see that since the time the genetic code was discovered about 1953, microscopes have grown very many times more powerful. The A on the left stands for an angstrom, a tenth of a nanometer (that is, a ten-billionth of a meter). 

Currently the most powerful microscopes can see things about 1 angstrom in width, which is a tenth of a nanometer. How does this compare to the sizes of the smallest units in brains? Those sizes are below:

Width of a neuron body (soma): about 100 microns (micrometers), which is about 1,000,000 angstroms.

Width of a synapse: about 500 nanometers, about 5000 angstroms.  When you search for "width of a synapse," you will commonly get a figure of 20 to 40 nanometers, but that is the width of the synaptic cleft, the gap between two synapses or between a synapse and a dendrite. The full head is much wider, as you can see from the page here.

Width of a dendritic spine: about 50 to 500 nanometers, about 500 to 5000 angstroms.

Length of a dendritic spine: the site here says, "the thin spine neck, which connects the spine to the main dendritic branch, has lengths between 0.04 and 1 μm [microns] and has 'door knob'-shaped head structures with diameters that between 0.5 and 2 μm [microns]." That length dimension is between 400 and 10,000 angstroms; and that head diameter is between 500 and 20,000 angstroms. 

The visual below (from the page here) shows an electron microscope image of a synapse. The width of the synaptic head is more than 500 nanometers (nm). We see nothing that looks like any kind of storage of human learned information. The neurotransmitters inside the spherical vesicles are short-lived chemicals that don't even last a week. 

Below we see a closeup electron microscope photograph of some dendritic spines which are 500 nanometers (5000 angstrom) wide, from the scientific paper here "Ultrastructural comparison of dendritic spine morphology preserved with cryo and chemical fixation," by Tamada et. al.  We see nothing that looks anything like stored learned information or stored memory information. Do a Google search for "diagram of dendritic spine head" and you will get diagrams that look like nothing that could be any system for storing information long-term. The diagrams will show that such dendritic spine heads are just bags of short-lived  proteins and chemicals. A few of these diagrams may show actin filaments looking a bit like a structure, but searching for "lifetime of actin filaments" you will be told that such filaments have lifetimes of only minutes. 

The only thing in the brain smaller than the structures shown above are protein molecules. But we know a reason why protein molecules cannot be a storage place for memories that can last for 50 years. The reason is that the average lifetime of a brain protein molecule is 1000 times shorter than the longest length that people can remember things.  Proteins in the brain have an average lifetime of two weeks or shorter. 

A scientific paper states this:

"Experience-dependent behavioral memories can last a lifetime, whereas even a long-lived protein or mRNA molecule has a half-life of around 24 hrs. Thus, the constituent molecules that subserve the maintenance of a memory will have completely turned over, i.e. have been broken down and resynthesized, over the course of about 1 week."

Research on the lifetime of synapse proteins is found in the June 2018 paper “Local and global influences on protein turnover in neurons and glia.” The paper starts out by noting that one earlier 2010 study found that the average half-life of brain proteins was about 9 days, and that a 2013 study found that the average half-life of brain proteins was about 5 days. The study then notes in Figure 3 that the average half-life of a synapse protein is only about 5 days, and that all of the main types of brain proteins (such as nucleus, mitochondrion, etc.) have half-lives of 15 days or less.  The 2018 study here precisely measured the lifetimes of more than 3000 brain proteins from all over the brain, and found not a single one with a lifetime of more than 75 days (figure 2 shows the average protein lifetime was only 11 days). 

The paper here states, "Experiments indicate in absence of activity average life times ranging from minutes for immature synapses to two months for mature ones with large weights."

Clearly the resolution of the most powerful microscopes is powerful enough to read memories stored in neurons or synapses or dendritic spines, if such memories existed. And more than 10,000 brains have been microscopically studied in recent years. The failure to microscopically read any  memories from human brain tissue is a major reason for thinking that brains do not store human memories. 

If memories were stored in the brain, roughly about the year 1960 humans would have been able to read learned information stored in brains, when the resolution of microscopes reached about 10 angstroms. If memories were stored in the brain, we would have discovered irrefutable evidence of such a thing about 60 years ago. The total failure to find a single speck of learned information in the brain by microscopic examination is one of the strongest reasons for disbelieving in a brain storage of memories. 

I predict with great confidence that microscopes will never find the slightest trace of learned information in the human brain, because memories are not stored in brains. But we may have in the future some occasional "false alarm" claims to have accomplished such a thing, claims that are not supported by robust evidence. Neuroscientists have often been guilty of both smoke-and-mirrors trickery and pareidolia, when someone claims to see something that isn't really there, typically because he is eagerly scanning large bodies of random, ambiguous data, eagerly hoping to find something that isn't really there. So we may see some pareidolia in which a neuroscientist claims to see a memory by microscopic examination. Such a thing will be like some fervent believer in animal ghosts in the clouds examining thousands of photos of clouds, and claiming that this one or that looks like the shape of an animal. 

DNA has merely low-level chemical information

Note that in the LiveScience article there is not any mention of any scientist sounding hopeful about a possibility of discovering stored information in the brain. 

This week I was reminded of the ability of the human mind to retain memories for 50 years, contrary to what we would expect from the high molecular turnover in brains.  I had a recollection which proved the ability of the mind to recall very old memories that have not been recalled in half a century. For some reason I recalled a book I had read about 50 years ago, and never since: the science fiction book "Galaxies Like Grains of Sand" by Brian Aldiss. I remembered some lines from the book. I wrote them down on paper like this:

"The mirror of the past lies shattered. The fragments you hold in your hand."

After I wrote this recollection of something I had not read, thought of or heard quoted in fifty years, I borrowed the book on www.archive.org.  I see that the lines were these (almost exactly as I remembered them)

"The long mirror of the past is shattered...Only a few fragments are left, and these you hold in your hand." 

For a person like me, the mirror of the past is not shattered, but remains well preserved after 60 years, contrary to what we would expect if memories were stored in brains with such high molecular turnover and such constant remodeling of synapses and dendritic spines. 

Below is a quote on the same topic from an earlier post discussing why brains cannot be the storage place of very old memories:

"I know for a fact that memories can persist for 50 years, without rehearsal. Recently I was trying to recall all kinds of details from my childhood, and recalled the names of persons I hadn't thought about for decades, as well as a Christmas incident I hadn't thought of for 50 years (I confirmed my recollection by asking my older brother about it). ...Upon looking through a list of old children shows from the 1960's, I saw the title “Lippy the Lion and Hardy Har Har,” which ran from 1962 to 1963 (and was not syndicated in repeats, to the best of my knowledge). I then immediately sung part of the melody of the very catchy theme song, which I hadn't heard in 53 years. I then looked up a clip on a youtube.com, and verified that my recall was exactly correct. I also recently recalled 'The Patty Duke Show' from the 1960's, a show I haven't seen in 50 years, and recalled that in the opening title sequence we saw Patty walking down some stairs. I looked up the title sequence on www.youtube.com, and verified that my 50-year-old memory was correct. This proves that a 53-year-old memory can be instantly recalled."

The prediction I make here is just one of several predictions in my 2019 post "Contrarian Predictions Regarding Biology, the Brain and Technology."  So far my predictions in that post are holding up very well. 

Why We Should Not Think the Human Brain Can Store Very Old Memories

Neurologists like to assume that all your memories are stored in your brain. But there are actually quite a few reasons for doubting this unproven assumption, including the research of scientists such as Karl Lashley and John Lorber. Their research showed that minds can be astonishing functional even when large parts of the brain are destroyed, either through disease or deliberate surgical removal. Lorber documented 600 cases of people with heavy brain damage (mostly due to hydraencephaly), and found that half of them had above average intelligence. Some children with brain problems sometimes undergo an operation called a hemispherectomy, in which half of their brain is removed. An article in Scientific American tells us, “Unbelievably, the surgery has no apparent effect on personality or memory.”

Given such very astonishing anomalies, we should give serious consideration to all arguments against the claim that your brain is storing all your memories. I will now discuss such an argument. The argument can be summarized as follows: there is no plausible mechanism by which the human brain could store very long-term memories such as 50-year-old memories. Every neurological memory theory that we have cannot explain any memories that have persisted for more than a year.

The Fact That Humans Can Remember Things for 50 Years

First, let's look at the basic fact of extreme long-term memory storage. It is a fact that humans can recall memories from 50 years ago. Some people have tried to suggest that perhaps human memory doesn't work for such a long time, and that remembering very old memories can be explained by the idea of what is called “rehearsal.” The idea is that perhaps a 60-year-old remembering is really just remembering previous recollections that he had at an earlier age. So perhaps, this idea goes, when you are 60 you are just remembering what you remembered from your childhood at 50, and that at 50 you were just remembered what you remembered from your childhood at age 40, and so forth.

But such an idea has been disproved by experiments. A scientific study by Harry Bahrick was entitled “Semantic memory content in permastore: Fifty years of memory for Spanish learned in school.” It showed that “large portions of the originally acquired information remain accessible for over 50 years in spite of the fact the information is not used or rehearsed.” The same researcher tested a large number of subjects to find out how well they could recall the faces of high school classmates, and found very substantial recall even with a group that had graduated 47 years ago. Bahrick reported the following:

Subjects are able to identify about 90% of the names and faces of the names of their classes at graduation. The visual information is retained virtually unimpaired for at least 35 years...Free-recall probability does not diminish over 50 yr for names of classmates assigned to one or more of the Relationship Categories A through F.

I know for a fact that memories can persist for 50 years, without rehearsal. Recently I was trying to recall all kinds of details from my childhood, and recalled the names of persons I hadn't thought about for decades, as well as a Christmas incident I hadn't thought of for 50 years (I confirmed my recollection by asking my older brother about it). Digging through my memories, I was able to recall the colors (gold and purple) of a gym uniform I wore, something I haven't thought about (nor seen in a photograph) for some 47 years. Upon looking through a list of old children shows from the 1960's, I saw the title “Lippy the Lion and Hardy Har Har,” which ran from 1962 to 1963 (and was not syndicated in repeats, to the best of my knowledge). I then immediately sung part of the melody of the very catchy theme song, which I hadn't heard in 53 years. I then looked up a clip on a youtube.com, and verified that my recall was exactly correct. I also recently recalled "The Patty Duke Show" from the 1960's, a show I haven't seen in 50 years, and recalled that in the opening title sequence we saw Patty walking down some stairs. I looked up the title sequence on www.youtube.com, and verified that my 50-year-old memory was correct. This proves that a 53-year-old memory can be instantly recalled.

So in trying to explain human memory, we need to have a theory that can explain human memories that persist for 50 years. Very confusingly, scientists use the term “long-term memory” for any memory lasting longer than an hour, which is very unfortunate because almost every thing you will find on the internet (seaching for “long term memory”) does not actually explain very long-term memory such as memories lasting for 50 years.

Why LTP and Synapse Plasticity Cannot Explain Very Long-Term Memory

Now let's look at neuroscientists' theories of memories. Quora.com is a “expert answer” website which claims to give “the best answer to any question.” One of its web pages asks the question, “How are memories stored and retrieved in the human brain?” The top answer (the one with most upvotes) is by Paul King, a computational neuroscientist. King very dogmatically gives us the following answer:

At the most basic level, memories are stored as microscopic chemical changes at the connection points between neurons in the brain..As information flows through the neural circuits and networks of the brain, the activity of the neurons causes the connection points to become stronger or weaker in response. The strengthening and weakening of the synapses (synaptic plasticity) is how the brain stores information. This mechanism behind this is called "long-term potentiation" or "LTP."

But there is actually no proof that any information is being stored when synapses are strengthened. From the mere fact that synapses may be strengthened when learning occurs, we are not entitled to deduce that information is being stored in synapses, for we also see blood vessels in the leg strengthen after repeated exercise, and that does not involve information storage. In order to actually prove that a synapse is storing information, you would need to do an experiment such as having one scientist store a symbol in an animal's brain (by training), and then have another scientist (unaware of what symbol had been stored) read that symbol from some synapses in the animal's brain, correctly identifying the symbol. No such experiment has ever been done.

The idea that a memory forms after repeating strengthening of synapses is inconsistent with the fact people very commonly remember things they experienced only one time. Almost every time someone tells you about a TV show they saw last night, or a conversation they recently had with a friend, they are remembering something they only were saw or heard one time, not through repeated experiences.

The evidence does not even clearly indicate that LTP correlates with memory, as one scientist's summary of experimental results indicates (a summary utterly inconsistent with the claim LTP is a general mechanism to explain memory).

What this means is that LTP and memory have been dissociated from each other in almost every conceivable fashion. LTP can be decreased and memory enhanced. Hippocampus-dependent memory deficits can occur with no discernable effect on LTP...There will be no direct quantitative or even qualitative relationship between LTP measured experimentally and memory measured experimentally—that is already abundantly clear from the available literature...The most damning observations probably are those examples where LTP is completely lost and there is no effect on hippocampus-dependent memory formation.

A scientific paper states this about LTP:

Based on the data reviewed here, it does not appear that the induction of LTP is a necessary or sufficient condition for the storage of new memories.

LTP is so weak an effect it is hard to even detect it. A scientific paper asks the following:

Why is it so difficult to see learning-associated synaptic changes?And does their absence in numerous experiments favor the null hypothesis?

What is misleadingly called “long-term potentiation” or LTP is a not-very-long-lasting effect by which certain types of high-frequency stimulation (such as stimulation by electrodes) produces an increase in synaptic strength. Synapses are gaps between nerve cells, gaps which neurotransmitters can jump over. The evidence that LTP even occurs when people remember things is not very strong, and in 1999 a scientist stated (after decades of research on LTP) the following:

[Scientists] have never been able to see it and actually correlate it with learning and memory. In other words, they've never been able to train an animal, look inside the brain, and see evidence that LTP occurred.

Since then a few studies have claimed to find evidence that LTP occurred during learning. But there is actually an insuperable problem in the idea that long-term potentiation could explain very long-term memories. The problem is that so-called long-term potentiation is actually a very short-term phenomenon. Speaking of long-term potentiation (LTP), and using the term “decays to baseline levels” (which means “disappears”), a scientific paper says the following:

Potentiation almost always decays to baseline levels within a week. These results suggest that while LTP is long-lasting, it does not correspond to the time course of a typical long-term memory. It is recognized that many memories do not last a life-time, but taking this point into consideration, we would then have to propose that LTP is only involved in the storage of short-term to intermediate memories. Again, we would be at a loss for a brain mechanism for the storage of a long-term memory.

A more recent scientific paper (published in 2013) says something similar, although it tells us even more strongly that so-called long-term potentiation (LTP) is really a very short-term affair. For it tells us that “in general LTP decays back to baseline within a few hours.” “Decays back to baseline” means the same as “vanishes.” 

Another 2013 paper agrees that so-called long-term potentiation is really very short-lived:

LTP always decays and usually does so rapidly. Its rate of decay is measured in hours or days (for review, see Abraham 2003). Even with extended “training,” a decay to baseline levels is observed within days to a week.

Scientists distinguish between two types of LTP: an E-LTP that can be produced by a single electrical stimulus, but only lasts one to three hours, and an L-LTP that requires multiple electrical stimulations, and can last about 8 hours or a little longer. But this fails to correspond to what we know about human memory, which is that humans can form a memory lasting decades after a single sensory experience. 

So evidently long-term potentiation cannot be any foundation or mechanism for long-term memories. This is the conclusion reached by the previous paper when it makes this conclusion about long-term potentiation (LTP):

In summary, if synaptic LTP is the mechanism of associative learning—and more generally, of memory—then it is disappointing that its properties explain neither the basic properties of associative learning nor the essential property of a memory mechanism. This dual failure contrasts instructively with the success of the hypothesis that DNA is the physical realization of the gene.

The book "Neuronal Mechanisms of Memory Formation" hints on page 451 that LTP may be a poor candidate for such a thing. It says this:

Definitive empirical support for synaptic plasticity modeled by LTP being a mechanism of memory processing is still lacking. For each piece of evidence that lends some support to the theory, there is likely to be equally strong evidence to suggest the contrary. The field of research reached a veritable stalemate some years ago when so-called cornerstones of research that supported the hypothesis were unable to be replicated...and the outcome was an increasing skepticism about whether LTP can be considered a neural substrate for learning and memory.

Referring to this scientific paper, another paper suggests that "LTP as a memory mechanism" may be more of a dogma than something well established by observations:

Shors and Matzel,,.concluded that LTP did not meet the criteria for providing a causal mechanism of memory. To make a long argument very short, they documented instances where changes in memory occur without LTP and where LTP occurs without changes in memory.....They report that between 1974 and 1997, more than 1300 articles occurred with “LTP” in the title. Of these, fewer than 80 described any behavioral manipulation relevant to assessing changes in memory. Furthermore, the articles that contained behavioral manipulations tended to provide evidence against the hypothesis that LTP is a memory mechanism. Thus, the claim that LTP is a molecular mechanism for learning and memory may be more of a dogma of neuroscientific memory research than a hypothesis that is being rigorously tested.  

A 2014 book stated, "Although LTP is considered to be the primary model for how learning and memory storage occur at the synapse level, the evidence supporting this claim is still inconclusive and speculative." A 1995 scientific paper found. "There is a striking negative correlation of spatial learning ability with LTP." This is the exact opposite of what we should expect if LTP was some type of memory mechanism. 

But what about syntaptic plasticity, previously mentioned in my quote from the neurologist King ? Since he claimed that LTP is the mechanism behind synaptic plasticity, and LTP cannot explain any memory lasting longer than a year, then synaptic plasticity will not work to explain very long-term memories.

To study LTP, scientists typically perform something artificial called tetanic stimulation, using a frequency of 100 hertz. A normal brain does not transmit signals at a frequency so high. Of the common brain waves (alpha, beta, and gamma), only gamma waves have a frequency of greater than 30 hertz, and such gamma waves generally do not have a frequency higher than 50 hertz.  LTP can be induced using a lower frequency, but only by prolonged stimulation such as stimulation lasting minutes. That type of sluggish low-frequency stimulation cannot explain human memories that can form instantly. 

Why Synapses Cannot Explain Very Long-Term Memory

Long-term memory cannot be stored in synapses, because synapses don't last long enough. Below is a quote from a scientific paper:

A quantitative value has been attached to the synaptic turnover rate by Stettler et al (2006), who examined the appearance and disappearance of axonal boutons in the intact visual cortex in monkeys.. and found the turnover rate to be 7% per week which would give the average synapse a lifetime of a little over 3 months.

You can read Stettler's paper here. A 2019 paper documents a 16-day examination of synapses, finding "the dataset contained n = 320 stable synapses, n = 163 eliminated synapses and n = 134 formed synapses."  That's about a 33% disappearance rate over a course of 16 days, suggesting an average synapse lifetime of less than three months.

You can google for “synaptic turnover rate” for more information. We cannot believe that synapses can store-long memories for 50 years if synapses only have an average lifetime of about 3 months. The paper here says the half-life of synapses is "from days to months."

Synapses often protrude out of bump-like structures on dendrites called dendritic spines. But those spines have lifetimes of less than 2 years.  Dendritic spines last no more than about a month in the hippocampus, and less than two years in the cortex. This study found that dendritic spines in the hippocampus last for only about 30 days. This study found that dendritic spines in the hippocampus have a turnover of about 40% each 4 days. This 2002 study found that a subgroup of dendritic spines in the cortex of mice brains (the more long-lasting subgroup) have a half-life of only 120 days. A paper on dendritic spines in the neocortex says, "Spines that appear and persist are rare." While a 2009 paper tried to insinuate a link between dendritic spines and memory, its data showed how unstable dendritic spines are.  Speaking of dendritic spines in the cortex, the paper found that "most daily formed spines have an average lifetime of ~1.5 days and a small fraction have an average lifetime of ~1–2 months," and told us that the fraction of dendritic spines lasting for more than a year was less than 1 percent. A 2018 paper has a graph showing a 5-day "survival fraction" of only about 30% for dendritic spines in the cortex.  A 2014 paper found that only 3% of new spines in the cortex persist for more than 22 days. Speaking of dendritic spines, a 2007 paper says, "Most spines that appear in adult animals are transient, and the addition of stable spines and synapses is rare." A 2016 paper found a dendritic spine turnover rate in the neocortex of 4% every 2 days. A 2018 paper found only about 30% of new and existing dendritic spines in the cortex remaining after 16 days (Figure 4 in the paper). 

Furthermore, it is known that the proteins existing between the two knobs of the synapse (the very proteins involved in synapse strengthening) are very short-lived, having average lifetimes of no more than a few days. A graduate student studying memory states it like this:

It’s long been thought that memories are maintained by the strengthening of synapses, but we know that the proteins involved in that strengthening are very unstable. They turn over on the scale of hours to, at most, a few days.

A scientific paper states the same thing:

Experience-dependent behavioral memories can last a lifetime, whereas even a long-lived protein or mRNA molecule has a half-life of around 24 hrs. Thus, the constituent molecules that subserve the maintenance of a memory will have completely turned over, i.e. have been broken down and resynthesized, over the course of about 1 week.

The paper cited above also states this (page 6):

The mutually opposing effects of LTP and LTD further add to the eventual disappearance of the memory maintained in the form of synaptic strengths. Successive events of LTP and LTD, occurring in diverse and unrelated contexts, counteract and overwrite each other and will, as time goes by, tend to obliterate old patterns of synaptic weights, covering them with layers of new ones. Once again, we are led to the conclusion that the pattern of synaptic strengths cannot be relied upon to preserve, for instance, childhood memories.

The latest and greatest research on the lifetime of synapse proteins is the June 2018 paper “Local and global influences on protein turnover in neurons and glia.” The paper starts out by noting that one earlier 2010 study found that the average half-life of brain proteins was about 9 days, and that a 2013 study found that the average half-life of brain proteins was about 5 days. The study then notes in Figure 3 that the average half-life of a synapse protein is only about 5 days, and that all of the main types of brain proteins (such as nucleus, mitochondrion, etc.) have half-lives of 15 days or less.  The 2018 study here precisely measured the lifetimes of more than 3000 brain proteins from all over the brain, and found not a single one with a lifetime of more than 75 days (figure 2 shows the average protein lifetime was only 11 days). 

The paper here states, "Experiments indicate in absence of activity average life times ranging from minutes for immature synapses to two months for mature ones with large weights."

When you think about synapses, visualize the edge of a seashore. Just as writing in the sand is a completely unstable way to store information, long-term information cannot be held in synapses. The proteins in between the synapses are turning over very rapidly (lasting no longer than about a week), and the entire synapse is replaced every few months.

In November 2014 UCLA professor David Glanzman and his colleagues published a scientific paper publishing research results. The authors said, “These results challenge the idea that stable synapses store long-term memories." Scientific American published an article on this research, an article entitled, “Memories May Not Live in Neuron's Synapses.” Glanzman stated, “Long-term memory is not stored at the synapse,” thereby contradicting decades of statements by neuroscientists who have dogmatically made unwarranted claims that long-term memory is stored in synapses.

Why Very Long-Term Memories Cannot Be Stored in the Cell Nucleus

His research has led Glanzman to a radical new idea: that memories are not stored in synapses, but in the nerve cell nucleus. In fact, in this TED talk Glanzman dogmatically declares this doctrine. At 15:34 in the talk, Glanzman says, “memories are stored in the cell nucleus – it is stored as changes in chromatin.” This is not at all what neurologists have been telling us for the past 20 years, and few other neuroscientists have supported such an idea.

We should be extremely suspicious and skeptical whenever scientists suddenly start giving some new answer to a fundamental answer, an answer completely different from the answer they have been dogmatically declaring for years. For example, if scientists were to suddenly start telling us that galaxies are not hold together by gravity (as they've been telling us for decades), but by, say, “dark energy pulsations,” we should be extremely skeptical that the new explanation is correct. In this case, there are very good reasons why Glanzman's recently-hatched answer to where long-term memories are stored cannot be right.

Chromatin is a term meaning DNA and surrounding histone protein molecules. Histone molecules are not suitable for storing very long-term memories because they are too short-lived. A scientific paper tells us that the half-life of histones in the brain is only about 223 days, meaning that every 223 days half of the histone molecules will be replaced.

So histone molecules are not a stable platform for storing very long-term memories. But what about DNA? The DNA molecule is stable. But there are several reasons why your DNA molecules cannot be storing your memories. The first reason is that your DNA molecules are already used for another purpose – the storing of genetic information used in making proteins. DNA molecules are like a book that already has its pages printed, not a book with empty pages that you can fill. The second reason is that DNA molecules use a bare bones “amino acid” language quite unsuitable for writing all the different types of human memories. The idea that somewhere your DNA has memory of your childhood summer vacations (expressed in an amino-acid language) is laughable.

The third reason is that the DNA of humans has been exhaustively analyzed by various multi-year projects such as the Human Genome Project and the ENCODE project, as well as various companies that specialize in personal analysis of the DNA of individual humans. Despite all of this huge investigation and analysis, no one has found any trace whatsoever of any type of real human memory (long-term or short-term) being stored in DNA. If you do a Google search for “can DNA store memories,” you will see various articles (most of them loosely-worded, speculative and exaggerating) that discuss various genetic effects (such as gene expression) that are not the same as an actual storage of a human memory. Such articles are typically written by people using the word “memories” in a very loose sense, not actually referring to memories in the precise sense of a recollection.

The fourth reason is that there is no known bodily mechanism by which lots of new information can be written to the storage area inside a DNA molecule. The fifth reason is that the DNA we see in brain neurons is basically identical to the DNA we see in other parts of the body (such as the DNA from foot cells). If memories were stored in DNA, the DNA in brain neurons would be much different from that of the DNA in other body parts. 

To completely defeat the idea that your memories may be stored in your DNA, I will merely remind the reader that DNA molecules are not read by brains – they are read by cells. It takes about 1 minute for a cell to read only the small part of the DNA needed to make a single protein (and DNA has recipes for thousands of proteins). If your memories were stored in DNA, it would take you hours to remember things that you can actually recall instantly. Thinking that DNA can store memories is like thinking that your refrigerator can cook a steak.

But couldn't very-long term memories just be stored in some unknown part of a neuron? No, because the proteins that make up neurons have short lifetimes. A scientist explains the timescales:

Protein half-lives in the cell range from about 2 minutes to about 20 hours, and half-lives of proteins typically are in the 2- to 4-hour time range. Okay, you say, that's fine for proteins, but what about "stable" things like the plasma membrane and the cytoskeleton? Neuronal membrane phospholipids turn over with half-lives in the minutes-to-hours range as well. The vast majority of actin microfilaments in dendritic spines of hippocampal pyramidal neurons turn over with astonishing rapidity—the average turnover time for an actin microfilament in a dendritic spine is 44 seconds...As a first approximation, the entirety of the functional components of your whole CNS [central nervous system] have been broken down and resynthesized over a 2-month time span. This should scare you. Your apparent stability as an individual is a perceptual illusion.

It is occasionally speculated that long-term memories might be stored in microtubules in a cell. But such things do not last long enough to be a storage place for memories lasting decades. A scientific paper tells us how short-lived brain microtubules are:

Neurons possess more stable microtubules compared to other cell types (Okabe and Hirokawa, 1988; Seitz-Tutter et al., 1988; Stepanova et al., 2003). These stable microtubules have half-lives of several hours and co-exist with dynamic microtubules with half-lives of several minutes.

Why Long-Term Memories Cannot Be Stored in the DNA Methylome

DNA methylation occurs when a very simple molecule becomes attached to part of a DNA molecule. Such a simple methyl molecule can act like a kind of flag that switches part of a gene on or off. The set of all of these methyl molecules attached to DNA is known as the DNA methylome. It has been suggested by a few that maybe memories are stored in this DNA methylome. 

The DNA methylome seems like a fairly stable thing, and so you don't have the “low stability” problem of very rapid protein molecule turnover that you have with the theory that memories are stored in synapses. But there are several reasons why it is not credible to maintain that human memories are being stored in such a DNA methylome.

The first reason is that we already know the function of this DNA methylome, that it is something other than storing memories. The methyl molecules that make up the methylome serve the purpose of genetic expression, a very different task than storing memories. If you were to maintain that the DNA methylome serves both of these purposes, it would be kind of like the Saturday Night Live comedic sketch that described a product like “Miracle Whip.” It went like this:

Wife: New Shimmer is a floor wax!
Husband: No, New Shimmer is a dessert topping!
Wife: It's a floor wax!
Husband: It's a dessert topping!
Wife: It's a floor wax, I'm telling you!
Husband: It's a dessert topping, you cow!
Spokesman: [ enters quickly ] Hey, hey, hey, calm down, you two. New Shimmer is both a floor wax and a dessert topping!

The second reason for doubting that memories are stored in the DNA methylome is that the DNA methylome couldn't be read with the speed needed for memory recall that is very fast. The DNA methylome consists of methyl molecules scattered across a DNA molecule. All evidence suggests that reading DNA is relatively slow. DNA transcription occurs at a rate of about 40 to 80 nucleotides per second, and there are billions of such nucleotides. But think of how fast people can recall memories. On the TV show Jeopardy we see people recalling very obscure memories in only a few seconds. When someone talks rapidly, he is retrieving language memories (such as the memory of what a particular word means) in a fraction of a second. That couldn't happen so fast if some relatively slow process of reading DNA was being used.

The third reason for doubting that memories are stored in the DNA methylome is that the methylome does not grow in size as learning occurs. As discussed here, the DNA methylome is larger (percentage-wise) in a newborn baby than in either a young adult or an old man.

The fourth reason for doubting that memories are stored in the DNA methylome is that methylation suppression experiments do not affect memory very dramatically. Scientists have ways of suppressing DNA methylation, and they have tested the effects of such suppression on learning and memory.  A scientific paper says that “inhibiting DNA methylation alters olfactory extinction but not acquisition learning.” Another scientific paper says that when DNA methylation was inhibited, “long-term memory strength itself was not affected.” 

A scientific paper states the following:

Prior studies found no effects of zeb, a DNA methylation inhibitor (Zhou et al., 2002), on learning in A. mellifera (Lockett et al., 2010; Biergans et al., 2012). More recently, Biergans et al. (2016), conducted a meta-analysis of multiple honey bee studies with methylation inhibitors (zeb and RG108), but found no strong overall effect of inhibiting DNA methylation on honey bee learning.

These are not the type of very dramatic effects on learning and memory that one would expect from DNA methylation inhibition if memories were being stored in the methylome. Other studies claiming a stronger effect of DNA methylation inhibition are typically unreliable studies using fewer than 15 animals per study group, and in such low-statistical-power studies there is a high chance of a false alarm or false positive.

There are actually two drugs for humans that work mainly by inhibiting DNA methlyation: azacitidine and decitabine.  If DNA methlyation was a mechanism for memory storage, we would expect that the side effect lists for these drugs would make some mention of a possible effect on learning or memory. But these drugs do not produce such a side effect. It has been claimed by the few proponents of memory stored in the DNA methylation marks that such marks are a stable medium for writing information. But search for "DNA methylation turnover" and you will find contrary claims, such as a paper entitled "Rapid turnover of DNA methylation in human cells." 

When It Comes to Explaining Very Long-Term Memory, Our Neuroscientists Are in Disarray

So how can we summarize the current state of scientific thought on how long-term memory is stored? The word that comes to mind is: disarray. In this matter our scientists are flailing about, wobbling this way and that way; but they aren't getting anywhere in terms of presenting a plausible answer as to how very long-term memory can be stored in the brain. Our scientists have done nothing to plausibly solve the permanence problem – the problem that very long-term memories cannot be explained by evoking transient “shifting sands” mechanisms such as LTP which last much less than a year (or in neurons, which are rebuilt every two months due to protein turnover). On this matter our scientists have merely presented explanatory facades – theories that do not hold up to scrutiny, like some movie studio building facade that you can see is a fake when you walk around and look behind it, finding no rooms behind the front.

Another sign of this disarray is a 2013 scientific paper with the title, “"Very long-term memories may be stored in the pattern of holes in the perineuronal net." After basically explaining in its first paragraph why current theories of long-term memories do not work and are not plausible, the author goes on to suggest a wildly imaginative and absurdly ornate speculation that perhaps the brain is a kind of a giant 3D punchcard, storing information like data used to be stored on the old 2D punchcards used by IBM electronic machinery in the 1970's. The author provides no good evidence for this wacky speculation, mainly discussing imaginary experiments that would lend support to it. The very appearance of such a paper is another sign that currently scientists have no good explanation for very long-term memory. I may note that IBM punchcards only worked because they were read by IBM punchcard-reader machines. In order for the brain to work as giant 3D punchcard, we would have to imagine a brain-reader machine that is nowhere to be found in the human body. There has never existed such a thing as a punchcard that can read itself.

This scientific article quotes a neuroscientist speculating about memory storage. The article says:

Neuroscience has also been struggling to find where the brain stores its memories. “They may be ‘hiding’ in high-dimensional cavities,” Markram speculates.

High-dimensional cavities? Cavities are holes, not information storage media. I think the quote bolsters my claim that scientists do not have any plausible explanation of how brains can be storing memories for 50 years. 

Often the modern neuroscientist will engage in pretentious talk which makes it sound as if there is some understanding of how very long-term memory storage can occur. But just occasionally we will get a little candor from our neuroscientists, such as when neuroscientist Sakina Palida admitted in 2015, “Up to this point, we still don’t understand how we maintain memories in our brains for up to our entire lifetimes.”

Concluding Remarks on Long-Term Memory

For the reasons given above, there is no plausible mechanism by which brains such as ours could be storing memories lasting longer than a year. There are only a few possible physical candidates for things that might store very long-term memory in our brain, and as we have seen, none of them are plausible candidates for a storage of very long-term memory.

The fact that our neurologists claim to have theories as to how very long-term memories could be stored does not mean that any such theory is tenable. Imagine if you lived on a planet in which your consciousness and long-term memory was due to a soul, and that the first time scientists dissected a brain, they found that the brain was filled with sawdust. No doubt such scientists would get busy inventing clever theories purporting to explain how sawdust can generate consciousness and long-term memories.

I may note that memories stretching back 50 years are inexplicable not merely from a neurological standpoint but also from a Darwinian standpoint. As I will argue later, from the standpoint of survival of the fittest and natural selection, there is no reason why any primate organism should ever need to remember anything for longer than about a year or two (it would work just fine to just keep remembering last year's memories). I may note that according to an article on wikipedia.com, the average life span in the Bronze Age was only 26 years old. There is no reason why natural selection (prior to the Bronze Age) would have equipped us to remember things for a length of time twice the average life span in the Bronze Age, and it is not plausible that very long-term memories are a recent evolutionary development.

In this post and in other places on this site, when I use the term “memory” I am referring to things such as episodic memories of life events, learned vocabulary, learned facts, and learned visual information that we can recall (such as identifying the name of an object, or recognizing a face). I do not refer to muscular skills such as the skill of how to ride a bike. Such skills are best referred to as “muscular skills” rather than memory.

Postscript: Richard Huganir led a study that was specifically dedicated to trying to find long-lived proteins in synapses. None were found.  The actual scientific paper found no such thing. Quite to the contrary, the paper found the following:

• Studying thousands of brain proteins, the study found that virtually all proteins in brains are very short-lived, with half-lives of less than a week.
  
• Table 2 of the paper gives specific half-life estimates for the most long-lasting brain proteins, and in this table only 10 out of thousands of brain proteins had half-lives of 10 days or longer.
  
• Of the proteins whose half-life is estimated in Table 2, only one of them has a half-life of longer than 30 days, that protein having a half-life of only 32 days.
  
• A graph in the paper indicates that none of the synapse proteins had a half-life of more than 35 days.
  

Below is a graph from the Huganir paper. It shows that the study found that virtually all proteins in synapses are very short-lived.

Below is another graph from the same paper. It shows that the study found that virtually all proteins in synapses are very short-lived.

Judging from these graphs, none of the proteins found had a half-life of longer than 35 days, and only a few had a half-life of more than 14 days.

The scientific paper (like very many scientific papers) had a misleading title that does not match its observations. The title was "Identification of long-lived synaptic proteins by proteomic analysis of synaptosome protein turnover." No such proteins were identified. All of the proteins studied had short lifetimes of less than 75 days; and almost all of the proteins studied had lifetimes of less than a week.

Post-postscript: In 2025 I had a recollection which proved the ability of the mind to recall very old memories that have not been recalled in several decades. For some reason I recalled a book I had read about 50 years ago, and never since: the science fiction book "Galaxies Like Grains of Sand" by Brian Aldiss. I remembered some lines from the book. I wrote them down on paper like this:

"The mirror of the past lies shattered. The fragments you hold in your hand."

After I wrote this recollection of something I had not read, thought of or heard quoted in fifty years, I borrowed the book on www.archive.org.  I see that the lines were these (almost exactly as I remembered them)

"The long mirror of the past is shattered...Only a few fragments are left, and these you hold in your hand." 

Post-post-postscript: The 2025 paper here states that  "the synaptic turnover rate is as high as 1% per day in the visual cortex." That is a rate of about 100% replacement every four months. If that is the typical rate at which synapes are replaced, synapses cannot be the storage place of memories lasting decades.  Quoting an even higher rate of synaptic turnover, the paper here states, "A recent imaging study revealed that the synaptic turnover rate in hippocampal CA1 cells is very high, with an estimated lifetime of 1–2 weeks (Attardo et al., 2015)."