US Government Gives Us Fake News About Brains and Memory

Courtesy of a sub-branch of the United States government, we have in today's science news an utterly bogus headline as phony as a three-dollar bill.  The headline is "Researchers uncover how the human brain separates, stores, and retrieves memories." The headline appears in a press release published by the National Institute of Neurological Disorders and Stroke, a branch of the National Institutes of Health (NIH), a branch of the US government.  

Scientists have no actual understanding of how memories form or how a human being is able to retrieve a memory. They have never been able to discover any credible coding mechanism or translation mechanism by which any of the main forms of human memories could be translated into neural states or synapse states. Computers have read-write heads to store information in particular places on a disk. The brain has nothing like a write component that could be used to store information in some particular part of the brain, and has nothing like a read component that could be used to read information from some particular part of the brain.  Computers have indexing systems and addressing systems that allow the instant retrieval of stored information. No such thing exists in the brain, which has no indexing system, no addressing, no coordinate system and no position notation system.  So the instant recall of a memory (given a single word or phrase) would seem to be impossible if such a recall occurs by the reading of neurons or synapses.  As discussed here, the extremely abundant levels of noise in the brain should make impossible both the accurate storage of learned information in the brai and the accurate retrieval of learned information from the brain.  And the many typically-overlooked slowing factors in the brain (such as synaptic delays) should make it impossible for a brain to be responsible for memory retrieval that can occur instantly.  Given the very short lifetimes of synaptic proteins (1000 times shorter than the longest length of time humans can remember things), and the high turnover of dendritic spines, no one has been able to come up with a credible theory of how brains could store memories that can last for 50 years.  Nor has any person been able to explain how the sluggish chemical operations in a brain could instantly form a memory, something humans routinely do. Learned memory information has never been discovered by examining any type of neural tissue. For example, not one single bit of a person's memory can be retrieved from a corpse or from some tissue extracted during brain surgery. 

The study in question ("Neurons detect cognitive boundaries to structure episodic memories in humans") involved 20 epilepsy patients who had electrodes planted in their heads, presumably for medical reasons such as determining the source of their seizures.  The patients were shown some videos, and some electrode readings were taken of electrical signals from their brain. In the press release we read the following:

"The researchers recorded the brain activity of participants as they watched the videos, and they noticed two distinct groups of cells that responded to different types of boundaries by increasing their activity. One group, called 'boundary cells' became more active in response to either a soft or hard boundary. A second group, referred to as 'event cells' responded only to hard boundaries. This led to the theory that the creation of a new memory occurs when there is a peak in the activity of both boundary and event cells, which is something that only occurs following a hard boundary."

I do not have access to the "Neurons detect cognitive boundaries to structure episodic memories in humans" paper, which is behind a paywall. But you can read for free the preprint of an identical-sounding paper by the same lead author (Jie Zheng) involving the same 20 epilepsy patients, the same claims, the same brain region (the medial temporal lobe), and the same experimental method involving taking electrode readings of brain signals while patients were watching videos.  That preprint ("Cognitive boundary signals in the human medial temporal lobe shape episodic memory representation") is not very impressive. 

The extremely dubious method followed was to arbitrarily select hundreds of neurons for study, and to look for some tiny subset of neurons with electrical activity that could be correlated (merely in some fraction-of-a-second blip way) with memory activity of the human subjects when "boundary conditions" of videos were shown, using the nickname "boundary cells" or "event cells" for such neurons.  The number of such "boundary cell" neurons found was reportedly 7%.  The first giant problem is that given many billions of neurons in the human brain, there is no reason to think that the arbitrarily selected set of hundreds of neurons had any involvement at all in the storage or retrieval of a human memory. In fact, there is a very strong reason for thinking that such neurons almost certainly would have had no involvement at all in the storage or retrieval of a human memory: the fact that a few hundred is such a tiny fraction of many billions. 

The second giant problem is that there is every reason to suspect that the small percentage of supposedly correlated neurons found (reportedly 7%) is just what we would expect to be finding by chance, when examining neurons with random electrical signals having nothing to do with memory.  The authors claim that chance would have produced a result of only 2% rather than 7%. But since the paper did not involve any blinding protocol (such as should have been used for a study like this to be worthy of our attention), we should not be impressed by such a difference.  We do not know whether the 7% is an over-estimate arising from scientists seeing what they wanted to see in a biased analysis occurring partially because of a failure to follow a blinding protocol which would have reduced analytic bias.  Also, we do not know whether the 2% is an under-estimate arising from scientists under-estimating things so that they could report a result that they wanted to report, in a biased analysis occurring partially because of a failure to follow a blinding protocol which would have reduced analytic bias.  

A similar state of affairs holds in regard to the report of the detection of cells calls "event cells."  The authors claim to have found that 6% of the studied hundreds of cells had some fraction-of-a-second correlation characteristic allowing them to be classified as "event cells," and they claim that only 2%  of cells would have such characteristics by chance.  But since the authors failed to follow any blinding protocol, we cannot have confidence in either of these numbers.  

Under the very unlikely scenario that some meaningful difference in neuron response has been detected here, there is no particular reason to think that it is some neural sign of memory formation or memory retrieval. There are any number of reasons why brain cells might respond differently while videos are being shown, most of which have nothing to do with learning or memory.  For example, a different visual stimulus can produce a different neural response, as can a different muscle movement or a fleeting emotion.  We are told that the "boundary conditions" in the watched videos (supposedly producing different responses in the so-called "boundary cells") were accompanied by "sharp visual input changes." So any difference in neural response might have been merely a difference related to different visual perceptions, not something having to do with memory. 

In short, no robust evidence has been provided in this preprint that any cells were involved in memory formation or memory retrieval, and since the "Neurons detect cognitive boundaries to structure episodic memories in humans" paper by the same lead author seemed to be identical in all the main features, there is no reason to think that such a study provided any evidence for a brain involvement in  memory formation or memory retrieval. 

Here is an excerpt from the press release touting the "Neurons detect cognitive boundaries to structure episodic memories in humans" paper, one that uses a faulty line of reasoning:

"The researchers next looked at memory retrieval and how this process relates to the firing of boundary and event cells. They theorized that the brain uses boundary peaks as markers for 'skimming' over past memories, much in the way the key photos are used to identify events. When the brain finds a firing pattern that looks familiar, it 'opens' that event.

Two different memory tests designed to study this theory were used. In the first, the participants were shown a series of still images and were asked whether they were from a scene in the film clips they just watched. Study participants were more likely to remember images that occurred soon after a hard or soft boundary, which is when a new 'photo' or 'event' would have been created.

The second test involved showing pairs of images taken from film clips that they had just watched. The participants were then asked which of the two images had appeared first. It turned out that they had a much harder time choosing the correct image if the two occurred on different sides of a hard boundary, possibly because they had been placed in different 'events.'

These findings provide a look into how the human brain creates, stores, and accesses memories." 

There is no justification for claiming that the experiments discussed in the quote above tell us anything about the brain. The experiments discussed in the quote above are psychology experiments involving only human mental performance, without any measurement of the brain.  What we see here is a trick that materialists frequently use:  use some experimental results that do not involve any brain reading or brain scanning or brain measurement, and then claim that such results tell you something about the brain. When experimental results merely tell us that humans perform in such-and-such a way, or merely tell us that minds perform in such-and-such a way, we have no warrant for saying that such results tell us that the brain is performing in such-and-such a way.

Not one single bit of robust evidence has been provided in the press release that any understanding has occurred as to how a brain could store or retrieve a memory, nor has any robust evidence been provided for the claim that brains store or retrieve memories.  All of the old reasons for rejecting such claims remain as strong as ever. 

In today's NIH press release we have an extremely untrue statement saying, "This work is transformative in how the researchers studied the way the human brain thinks." No, the study described is just another example of a dubious neuroscience research design like I have seen countless times before.  The study was funded by the NIH's Brain Initiative, and the PR people of that project have before often groundlessly used the word "transformative" for meager research results.  I quote from a previous post of mine discussing the lack of major progress made by the Brain Initiative:

"So far the BRAIN Initiative has been running for four or five years, and has accomplished nothing extremely noteworthy. Our understanding of the brain has not dramatically advanced during those four or five years, and all the old mysteries of mind and memory seem as mysterious as ever. At this 'Achievements' link there is a discussion of what the BRAIN Initiative has accomplished so far. At the top of the text is a big bold headline saying 'Transformative Advances,' but the BRAIN Initiative has produced no such transformative advances. Go beyond the flashy spin on the web site, the high-tech glitter, and the discussion of things in progress that haven't yet yielded much, and you have not a single major accomplishment relating to our understanding of the mind or memory. You see in this section a video entitled 'The BRAIN Initiative – the First Five Years.' The video fails to list a single accomplishment of the BRAIN Initiative. Apparently all this work to mechanistically explain the mind is pretty much a flop and a failure so far."

Below is an extremely relevant quote from the well-worth-reading paper "A Call for Greater Modesty in Psychology and Cognitive Neuroscience":

"A romantic view holds that science is built on different values, such as integrity and honesty, as well as different systems of operation that mandate a dispassionate, calculated and systematic pursuit of the 'truth'. However, such a view of science is naïve. The incentive structure of modern science is such that a 'simplify, then exaggerate' strategy has become dominant, even if only tacitly. To get published in leading journals, to be awarded grants and to be hired as a postdoc or faculty member, a system-wide bias for novelty, exaggeration and storytelling has emerged (Huber et al., 2019; Nosek et al., 2012). The prizing of novelty over quality represents one overarching driver in the construction of a research culture beset by the widespread use of questionable research practices and low levels of reproducibility (Chambers, 2017; Munafò et al., 2017; Nelson et al., 2018; Open Science Collaboration, 2015; Simmons et al., 2011). Indeed, although there have arguably been recent successes (Shiffrin et al., 2018), many aspects of modern psychology and brain science resemble a creative writing class as much as a systematic science of brain or mind."