Sunday, May 28, 2023

How Neuroscientists and Science Writers Try to Fool You Into Thinking Brains Store Memories

There is no robust evidence that brains store human memories. But neuroscientists and science writers have a series of tricks they use to try to make you think that there is evidence that brains store memories. Below is a list of some of these tricks. 

Trick #1: Neuroscientists teach some animals something, observe some randomly selected synapses or neurons, and then claim evidence of "experience-dependent synaptic plasticity" or "learning-dependent brain changes."

This trick can be  implemented by performing steps such as these:

(1) Observe brain cells and synapses in various randomly selected tiny parts of the brains, perhaps 10, 20 or 30 different areas that we may call "the tiny study areas." 

(2) Teach some animals (usually rodents) something, typically over a time period of several days or weeks. The easiest way to do is by fear conditioning. Animals can be put in a cage, with a shock plate that delivers a jolt of electricity whenever the animals step on it.  So the thing being taught is simply: avoid the little shock plate, for that will produce pain. 

(3) Observe again the same tiny study areas, and look for changes or differences.

(4) Write up a description using whichever tiny study area seemed to show the biggest difference after the learning occurred. Claim that such an area shows evidence of "experience-dependent synaptic plasticity" or "learning-dependent brain changes."

There is a giant reason why this technique is illegitimate. The reason is that all synapses of the brain undergo constant change regardless of whether learning has occurred.  This change is sometimes called synaptic turnover or spontaneous synaptic remodeling. So showing that some change occurred while learning occurred does nothing to show that learning produced such a change.

A scientific paper makes it clear that synaptic remodeling not caused by learning is constantly occurring. It states the following:

"Synaptic remodeling is driven by both activity-dependent and spontaneous processes. The magnitude of spontaneous synaptic remodeling is comparable with that of activity-dependent remodeling. Spontaneous synaptic remodeling processes can give rise to a full repertoire of synaptic sizes even in the complete absence of activity... Spontaneous remodeling processes drive continuous fluctuations in the properties of individual synapses." 

We know why such spontaneous fluctuations and instability occurs. It is largely because the protein molecules that make up synapses are shorted-lived, having average lifetimes of less than two weeks. Such molecules are constantly being replaced, causing instability and remodeling.  Synapses are like vines between trees in the Amazon rain forest, showing the same ever-changing nature, regardless of whether learning has occurred. 

The paper above tells us this:"Imaging studies in primary culture indicate that synaptic configurations erode significantly over timescales of a few days." We also read this:

"These observations – that synapses of all sizes and types form at normal proportions – have profound implications: when synapses and dendritic spines are initially formed, they generally have small volumes and PSDs..Their subsequent conversion into large, mushroom-shaped spines is often thought to be driven by activity-dependent potentiation...However, as mentioned above, large synapses, including mushroom-shaped spines, develop in normal proportions even when activity is essentially nonexistent... It thus seems that activity-independent processes can generate the full repertoire of synaptic sizes, highlighting the potency of spontaneous synaptic remodeling."

It seems that synapses are rather like clouds: configurations of matter that do not maintain the same structure over long periods of time. There is a reason why such ever-changing, physically unstable behavior in synapses is very important. The first reason is that such very great instability in the structure of synapses means that they cannot be the storage place of memories that can last unchanged for fifty years or more. The second reason is that such behavior in synapses invalidates the type of evidence typically given for synaptic memory storage. 

If you think that synapses normally keep their structure unchanged for years, then you may think it means something if a scientist observes some synaptic change while some organism learned something over the course of several days. You may think that such synapses were "molded" or shaped by such learning. But if you correctly understand that synapses are things as unstable as the wet sand at the edge of a seashore or the leaves of a maple tree, then you should not think that it means anything if a scientist observes some synaptic change while some organism learned something over the course of several days -- because there is no way to tell whether such changes occurred because of the learning.   


Scenario 1: Synapses are stable, retaining their size and configuration for very many years, being changed only by learning or experience.

Scenario 2: Synapses all over the brain are unstable, subject to constant remodeling and fluctuations, like the leaves of maple trees that grow in the spring and fall off in autumn, or like ever-changing clouds or shifting snow drifts, or like vines in the Amazon rain forest that frequently change in size and appearance.

                            CONSEQUENCES

Synapses might conceivably store memories that remain stable for 50 years or more.

Synapses cannot possibly store memories that remain stable for 50 years or more.

Observing some synaptic change during learning might be evidence that such change occurred because of the learning.

Observing some synaptic change that occurred during learning is no evidence that such change occurred because of the learning. Such change was probably caused by the synaptic remodeling and fluctuations that constantly occurs all over the brain, regardless of whether any learning occurs.

Scenario 2 above is the correct scenario. So it is fallacious to teach some animals something, observe some randomly selected synapses or neurons, and then claim evidence of "experience-dependent synaptic plasticity" or "learning-dependent brain changes." Such a procedure does nothing to show that the synapse changes were caused by the learning that occurred. 

Trick #2: Neuroscientists discuss technical details of changes in synapses, and then try to pass off such changes as knowledge of memory storage, even though there is no evidence synapse changes involve memory storage.

When this trick is used, a neuroscientist tries to impress us by giving many chemistry details and biology details. We may hear an in-depth discussion of the exact chemistry going on during some change in a synapse. Or we hear an in-depth discussion of some chemistry called LTP. Or we may hear an in-depth discussion of what occurs when a chemical signal jumps across a synaptic gap. This will be described as a discussion of "how memory storage occurs." 

But it is always very dubious or fallacious to be describing such details as a description of how memory storage occurs. We have no evidence that such biochemical activity actually involves any memory storage. There are also extremely strong reasons for disbelieving that such chemical activity is any such thing as a storage of memories. For example, we know that the proteins that make up synapses have very short lifetimes of two weeks or less, and we know that synapses are subject to constant random fluctuations or remodeling. But memories can last in humans for 50 years or longer. So no one should be describing synapse behavior and calling that discussion a discussion of memory storage. Similarly, we know that so-called LTP (a term meaning "long-term potentiation") is actually a very short-lived effect that does not last longer than a year. So no one should be describing LTP and calling that a discussion of what we know about memory storage. 

An example of Trick #2 is the extremely misleading paper "Understanding the physical basis of memory: Molecular mechanisms of the engram." No actual evidence is given in this paper that any of the authors have any such thing as an understanding of a physical basis of memory.  The authors mainly just discuss low-level chemistry, and call that (without justification) a description of memory storage. The authors try to maximize their use of the word "engram," a speculative word that no scientist should ever be using in a matter-of-fact way. An engram is an alleged cell or cells storing  a memory. No such thing has ever been detected.  When authors such as this go about using the word "engram" in every possible opportunity, they are like cosmologists who litter their papers with matter-of-fact references to "dark matter" and "dark energy," two things that have never been detected. 

Towards the end of the paper, the authors give away that they have no real knowledge of a neural storage of memory, by asking a long series of questions showing that they have no real understanding of memory storage, memory encoding or memory retrieval. 

Trick #3: Neuroscientists electrically stimulate some part of an organism's brain, observing "freezing" behavior, and then claim that they have artificially evoked a brain-stored memory by doing such electrical stimulation. 

To understand how this trick works, I must explain a very defective way in which neuroscientists often attempt to get a sign of memory recall in animals.  It works like this:

(1) A rodent is placed in a cage that has a shock plate, and if the animal steps on the shock plate, it receives a painful jolt of electricity. 

(2) Later the same animal is placed in a cage that either has such a shock plate, or maybe a different type of shock plate, or maybe no shock plate at all. Some attempt is made to judge what percentage of the time in the cage the animal is motionless. Such a percentage is called a "freezing percentage." The more immobile the animal is in the cage, the more it is claimed that the animal was recalling a memory of receiving a painful shock from the shock plate. The authors try to insinuate that if the rodent is more immobile, that means the rodent was "freezing" from fear. 

This is all a very poor and inaccurate way to measure fear in animals. If you don't use software to measure such a thing, judging what percentage of the time the animal was immobile will require a subjective judgment. Even if you do use "freezing measurement software," the resulting number will depend on subjective software parameters that can be varied from session to session. Freezing (non-movement) is not a standard way that animals react when faced with something producing fear. It is just as likely that animals will try to flee, and it is hard to disentangle freezing behavior from fleeing behavior.  

There is a reliable to measure fear in rodents: you can measure heart rate, which spikes very dramatically in rodents when they are afraid. But foolishly neuroscientists have a custom of measuring fear by using unreliable "freezing" judgments rather than reliable heart rate measurements.   

When Trick #3 occurs, what goes on is this:

(1) Rodents will be fear-trained using shock plates in cages, plates that give a painful zap to the rodent's feet.

(2) Some guess will be made about some place in the rodent's brain where the memory of this experience may be stored.

(3) That place will be electrically stimulated or stimulated using optogenetics, something rather like a laser.

(4) Not very accurate "freezing behavior" methods will be used to try to detect whether the rodent recalled the fearful memory after the electrical stimulation or optogenetic stimulation of part of the rodent's brain. If the "freezing percentage" number is higher than the number produced when the animal's brain is not being electrically stimulated or stimulated by optogenetics, this will be cited as evidence that the fearful memory was recalled by the animal, and that the memory was brain-stored. 

It is easy to explain why such a method does nothing to show that a fearful memory is actually being recalled.  So called "freezing behavior" is judged to have occurred whenever a rodent is immobile in a cage. What percentage of the time in a cage an animal will be immobile is a random parameter that will randomly vary from one experimental session to another. So with any experiment using the procedure above, you would have a 50% chance of getting a "higher freezing percentage" even if the brain stimulation does nothing to produce any recall of a fearful memory, and even if no fearful memory is being recalled. So getting a higher "freezing percentage" after brain stimulation does nothing to show that a fearful memory was recalled, and nothing to show that memory is stored in brains. 

In fact, there is strong reason to suspect that the very act of stimulating a brain with electricity or optogenetic stimulation may itself produce "freezing behavior," making such a technique all the more worthless for showing that a memory was stored in a brain or that recall of a memory occurred because of artificial stimulation.   A science paper says that it is possible to induce freezing in rodents by stimulating a wide variety of brain regions. It says, "It is possible to induce freezing by activating a variety of brain areas and projections, including the hippocampus (Liu et al., 2012), lateral, basal and central amygdala (Ciocchi et al., 2010); Johansen et al., 2010; Gore et al., 2015a), periaqueductal gray (Tovote et al., 2016), motor and primary sensory cortices (Kass et al., 2013), prefrontal projections (Rajasethupathy et al., 2015) and retrosplenial cortex (Cowansage et al., 2014).” 

Heart rate dramatically increases in rodents when they are afraid. The only reliable way to detect whether a rodent is afraid is to measure very sharp heart rate spikes. Any rodent experiment that fails to use such a technique not only fails to show that memories are stored in brains, but also fails to even show that a fearful memory is being recalled. 

Trick #4: Neuroscientists and science writers make spurious claims that "engrams"or "engram cells" have been discovered, and attempt to maximize use of the never-justified word "engram."

An example of Trick #4 is the extremely misleading paper "Understanding the physical basis of memory: Molecular mechanisms of the engram." No actual evidence is given in this paper that any of the authors have any such thing as an understanding of a physical basis of memory.  The authors try to maximize their use of the word "engram," a speculative word that no scientist should ever be using in a matter-of-fact way. An engram is an alleged cell or cells storing  a memory. No such thing has ever been detected.  When authors such as this go about using the word "engram" in every possible opportunity, they are like supersymmetry theorists who litter their papers with matter-of-fact references to "superpartner particles," things that have never been observed.  For example, the authors make this very untrue statement:

"We now know that learning activates a group of neurons and induces on them a set of physical changes to form the engram. During learning, these engram cells exhibit plasticity of their synaptic strengths—the connections between them get strengthened."

No one knows any such thing. "Synapse strengthening" is a vacuous phrase that does nothing to explain how memories could be written to synapses. Information is written through a writing of symbolic tokens, never by a mere strengthening.  Synapse strengthening is a very slow process, requiring many minutes. The claim that memories are written when synapses are strengthened is inconsistent with the fact that humans can instantly form new memories, forming much, much faster than the slow pace of synapse strengthening.  Moreover, the short lifetimes of synapse proteins and the instability of synapses means that synapses cannot be the storage place of memories. 

In his Nautilus post “Here's Why Most Neuroscientists Are Wrong About the Brain,” C. R. Gallistel (a professor of psychology and cognitive neuroscience) points out the absurdity of thinking that mere changes in synapse strengths could store the complex information humans remember. Gallistel writes the following:

"It does not make sense to say that something stores information but cannot store numbers. Neuroscientists have not come to terms with this truth. I have repeatedly asked roomfuls of my colleagues, first, whether they believe that the brain stores information by changing synaptic connections—they all say, yes—and then how the brain might store a number in an altered pattern of synaptic connections. They are stumped, or refuse to answer....When I asked how one could store numbers in synapses, several became angry or diverted the discussion with questions like, 'What’s a number?' ”

What Gallistel describes sounds dysfunctional: a pretentious neuroscientist community that claims to understand how memory can be stored in a brain, but cannot give anything like a plausible answer to basic questions such as “How could a number be stored in a brain?” or “How could a series of words be stored in a brain?” or “How could a remembered image be stored in a brain?” Anyone who cannot suggest plausible detailed answers to such questions has no business claiming to understand how a brain could store a memory, and also has no business claiming that a brain does store episodic or conceptual memories.   

Trick #5: Neuroscientists do badly designed junk science experiments involving too small study group sizes, a lack of blinding, and a lack of pre-registration.

(1) Scientists doing experiments involving memory typically use study group sizes that are too small to produce any reliable result. The results are mainly false alarms of a type that can easily arise when too-small study group sizes are used. 
(2) Scientists doing experiments involving memory typically fail to do the sample size calculations that would alert them that the study group sizes they are using are way too small to produce a reliable result. 
(3) Scientists doing experiments involving memory are very often using defective experimental procedures that produce unreliable results, such as trying to measure fear in rodents by subjectively judging "freezing behavior" rather than using better procedures producing more reliable results, such as trying to measure fear in rodents by measuring heart rate (which reliably spikes very sharply when a rodent is afraid). 
(4) Scientists doing experiments involving memory routinely fail to follow a blinding protocol that would reduce the chance of them producing false-alarm results in which they merely "see what they want to see." 
(5)  Scientists doing experiments involving memory routinely fail to follow good practices by pre-registering an exact experimental method for collecting and analyzing data. Often their papers show strong signs of "keep torturing the data until it confesses," which can also be described as "keep slicing and dicing the data until you find something like you hoped to find." 

Trick #6: Neuroscientists zap human brains with electricity or lasers in small sample studies of memory recall, and use subset mining or other tricks to make it look like some improvement in recall occurred. 

Neuroscientists sometimes do studies that involve zapping the brains of subjects while they are trying to memorize or recall, and neuroscientists sometimes claim this produces superior memory recall. Invariably the results can be best explained by simple ideas that do not involve any physical effect by which memory is improved. Usually the results are simply results that are not very improbable, given random variations in test results.  Typically the study group sizes will be very small. It is all too easy to produce all kinds of false alarms when you are dealing with small study group sizes, the type of effects that will disappear when larger study group sizes are used. Often what goes on in such studies is subset mining. Testing 20 subjects, the result will be that the overall performance no better than a control group that received no electrical stimulation. But then the neuroscientists will try slicing and dicing the data until some "statistically significant" result can be claimed.  

For example, they may look for some better performance on the most recent words studied, or the middle words studied, or the last words studied. Or they will look for some better performance on the shortest words, or the longest words, or the most tangible-sounding words. Or, if they stimulated 20 different regions of the brain, they will look for better results with one of these 20 regions. If you slice and dice the data enough, it is easy to find some little slice that shows a better-than-average result, and then claim that as your result.  Such results show nothing.  You could get the same results if you had the same number of subjects rubbing a rabbit's foot while trying to memorize words. 

We are typically told in such experiments that there was a "sham" control group that had electrodes attached to its head, but received no brain stimulation, which was compared to a group that received real brain stimulation from electrodes. But a paper tells us that when brain stimulation occurs, "The subject feels, sees, or hears something and thereby knows stimulation has occurred.And the article here refers to one of these brain stimulation experiments and says, "People reported feeling things like itching, tingling, poking and warming as the device ramps up or down, for the first and last 30-60 seconds of treatment, Reinhart said."

We can imagine how things would go if subjects are told that they will be part of either a group getting real brain simulation, or a "sham" group getting no such stimulation. The subjects recognizing they are in the "real stimulation" group would tend to feel an obligation to perform better, while those recognizing they are in the "sham" group would feel no such obligation, and perhaps think they are expected to perform worse. This alone could account for the kind of slight difference typically reported in results between "sham" groups getting no brain stimulation and "real" groups actually getting stimulation, without the evidence being any evidence of brain memory storage. You could probably get the same difference in memory scores by using a low voltage cattle prod on someone's buttocks. While people were being zapped on their buttocks, they might think, "Damn, I better hurry up and finish this memorization task!" That alone could account for any difference in performance. 

Press releases about such studies are typically misleading.  One press release mentions a study of "150 people," failing to tell us that none of the study group sizes were larger than 20 (too small a study group size for a reliable result). We hear of some big memory improvement not justified by any results in the corresponding paper, which shows results consistent with chance fluctuations.  

The idea that memory recall can be improved by zapping a brain with electricity is a silly one. Under the theory that memories are stored in brains, we would expect that electrical stimulation would tend to disrupt memories, not improve recall. 

What's going on these days is that neuroscientists and their press flunkeys are resorting to the trick of trying to cover up the ridiculously inadequate study group sizes they are using. The trick works like this: you claim that some study involved a certain number of subjects, even though each of the experiments only involved a small fraction of that number.  An example is found in a study announced with the phony headline "Short Term Memory Problems Can Be Improved With Laser Therapy." The nonsensical idea is that you can improve people's memory by zapping their heads with lasers. The press release announces that "researchers at Beijing Normal University carried out experiments with 90 male and female participants aged between 18 and 25."  But the scientific paper tells us this, which is a minor masterpiece of misleading wording:

"Neurologically normal college students (n = 90) with normal or corrected-to-normal vision participated in four experiments. Of these, 27 participated in experiment 1 (five males, mean age = 22). No statistical methods were used to predetermine the sample size, but the sample size was chosen to be adequate to obtain robust results as determined by preliminary experiments. Because the identified tPBM effect in experiment 1 was robust, we set the sample size to 21 in experiments 2 to 4 (experiment 2: seven males, age range = 22.8 ± 3.8; experiment 3: eight males, age range = 22.7 ± 4.1; experiment 4: seven males, age range = 22.8 ± 4.0). Data from 11 participants (four in experiment 1, three in experiment 2, two in experiment 3, and two in experiment 4) were excluded because of incomplete data or low EEG quality."

Here is why the above can be called "a minor masterpiece of misleading wording":
(1) None of the experiments used more than 27 subjects, but the casual reader may get the impression that each used 90 subjects.
(2) The casual reader will get the impression that each of experiments 2, 3, and 4 involved 21 subjects, although a very close reading will show that none of them involved more than 8 subjects (21 being the rough total of seven subjects used in experiment 2, 8 subjects used in experiment 3, and seven subjects used in experiment 4). 

It is obvious why "no statistical methods were used to predetermine the sample size": because the ridiculously small study group sizes used were way too small for a reliable result.  The result is almost certainly a mere false alarm "noise" result, and provides no evidence that zapping people's heads with lasers can improve their short-term memory. The impression given by the press release that 90 subjects were used for these experiments is very misleading, since most of them used fewer than 10 subjects. 

Trick #7: Neuroscientists and science writers cite badly designed junk science experiments involving too small study group sizes, a lack of blinding, and a lack of pre-registration, repeating the never-justified claims of such studies.

This trick is very easy to do. A neuroscientist can simply cite other studies claiming results that the neuroscientist wants you to believe in, without paying any attention to whether such studies were poorly designed and produced robust results. This type of thing goes on massively in neuroscience papers. Junk poorly designed experimental studies are produced in vast numbers, and such studies are cited in vast numbers.  

I can give an example. The paper "DNA Methylation and Establishing Memory" tries to get you believe in the groundless that DNA methylation has something to do with memory storage. A key part of the paper is a section entitled "DNA methylation alterations upon the creation of long-term memories." There we read this (most of which are groundless claims):

"This is referred to as contextual fear conditioning. 46 This can result in a life-long fearful memory after a single training event. 46 While the long term memory of this event appears to be first stored in the hippocampus, this storage is transient and does not remain in the hippocampus. 46 Much of the long term storage of contextual fear conditioning memory appears to take place in the anterior cingulate cortex (Figure 2). 55 Contextual fear conditioning applied to a rat induces more than 5000 differentially methylated regions (of 500 nucleotides each) in the rat hippocampal neuronal genome as measured both 1 and 24 hours after the conditioning. 4 This causes about 500 genes to be up-regulated (possibly due to hypomethylation of CpG sites) and about 1000 genes to be down-regulated (observed to be correlated with newly formed 5-methylcytosine at CpG sites in promoter regions). Overall, 9.17% of the genes in the genomes of rat hippocampus neurons are differentially methylated after the conditioning event. The pattern of induced and repressed genes within neurons appears to provide a molecular basis for forming this first transient memory of this training event in the hippocampus of the rat brain. 4 When similar contextual fear conditioning is applied to a mouse, 1 hour after contextual fear conditioning there were 675 demethylated genes and 613 hypermethylated genes in the hippocampus region of the mouse brain. 3 These changes were transient in the hippocampal neurons, and almost none were present in the hippocampus after 4 weeks. However, in mice subjected to contextual fear conditioning, after 4 weeks there were more than 1000 differentially methylated genes and more than 1000 differentially expressed genes in the anterior cingulate cortex, 3 where long-term memories are stored in the mouse brain. 55

We have in the references above no references to any well-designed and solid experimental studies. The references include the following:

Reference 4: A reference to the poorly designed study described in the  science paper  "Experience-dependent epigenomic reorganization in the hippocampus.” This is a study used way too small study group sizes such as groups of only three rats or four rats. The study never mentions any adequate study group size. 

Reference 3: A paywalled paper.

Reference 55: A reference to the poorly designed study described in the  science paper  "The Involvement of the Anterior Cingulate Cortex in Remote Contextual Fear Memory."  This is another paper with way too small study groups sizes such as only 6 animals. 

Reference 46: The paper cited provides no evidence that memories are stored in the hippocampus. 

Trick #8: They refer to mere observations of brain tissue, making unjustified claims such as "Scientists watch a memory being formed" or "Scientists show how memories are formed in brains."

An example was that when  there appeared a scientific paper merely claiming that "Regional synapse gain and loss accompany memory formation in larval zebrafish," there appeared a great number of press stories repeating the headline of a press release claiming that the formation of a memory had been observed (a claim not made in the paper).  We have every reason to believe that synapse gains and losses occur  continually in the human body, regardless of whether some new memory is forming.  The study above looked for synapse gains and synapse losses in zebrafish, finding an equal number of losses and gains. No evidence was provided that the zebrafish that learned better had gained more synapses than those that learned less. The study was announced with untrue press stories making claims that scientists had learned how memories form. 

Trick #9: They do some software project claimed as a "brain simulation," and try to fool you into thinking that this sheds light on how a brain could store human memories.

So-called "neural nets" are computer software systems that are said to be inspired by the brain, but which have little actual similarity to human brains. Such "neural nets" use mainly a technique called backpropagation, but it is not believed that the human brain uses any such technique. 

In backpropagation software nodes receive inputs from multiple other nodes, and then act as inputs to other nodes, with inputs traveling forwards and backwards. The diagram below gives you a very rough idea of what is going on:

network

Referring to backpropagation (also called backprop), a 2020 paper in Nature Neuroscience entitled "Backpropagation in the Brain" says, "There is no direct evidence that the brain uses a backprop-like algorithm for learning." In a speculative science document entitled "How to do backpropagation in a brain," we read, "Many neuroscientists think back-propagation is biologically implausible because they cannot see how neurons could possibly do it." At the scientific web site Quanta we read this: "For a variety of reasons, backpropagation isn’t compatible with the brain’s anatomy and physiology, particularly in the cortex." The article quotes the famous biologist Francis Crick: "As far as the learning process is concerned, it is unlikely that the brain actually uses back propagation.” The article mentions some other methods a little like backpropagation that could possible by used by the brain, but states, "Nevertheless, concrete empirical evidence that living brains use these plausible mechanisms remains elusive.

A software project using some kind of neural net might somehow be able to store something like learned information through some kind of "weight modification" deal, but that does nothing to explain how a brain could store a memory, or show that a brain could store a memory. So-called "neural nets" are not realistic simulations of the brain. One gigantic difference is that in a neural net information travels from node to node with perfect fidelity, but in the human brain whenever a signal passes across a synaptic gap, the signal transmission occurs with less than 50% reliability.  Another gigantic difference is that neural nets are stable, while synapses are very unstable because of their short-lived proteins. So showing that some information can be stored in a so-called neural net does nothing to show that a brain can store memories. 

Trick #10: They do or discuss some psychology experiment on memory, and try to pass that off as evidence for brains storing memories

You might call this the "psychology results dressed up as neuroscience results" trick. When this trick is done,  the discussion will be centered around some psychology experiment involving memory.  The data from the experiment (which may be good, solid data) will be mixed up with claims of neural involvement. For example, some study may show that people learn less when they are distracted; and this may be described as a study showing that the brain forms memories less effectively when people are distracted. The discussion of this experiment may be speckled with speculation about how a brain might store a memory, or which parts of the brain may be involved in a storage of memory. In such cases you have to ask yourself: has anything been done here to show that brains stored memories, or did the experiment merely show that people can acquire new memories, and learn new things? It is an undisputed fact that humans and animals can learn new things and acquire new memories. Some study merely showing that humans or animals learned things or remembered things does nothing to show that a brain stored a memory. Various combinations occur. A study may involve nothing but psychology research (described dogmatically as telling something about the brain), or in other cases the study may be something like 85% psychology research with 15% brain research, with the brain research being some not-very-relevant or weakly established stuff added so that the paper can be passed off as neuroscience research rather than psychology research.  

Some ways to figure out when you've been tricked

Detecting the tricks of neuroscientists can be tricky. One way to do that is to ask some basic questions such as these:

(1) Has the neuroscientist stated any credible theory of how experiences or learned information could be translated into neural states?

(2) Has the neuroscientist stated any credible theory of how memories could be instantly formed in a brain?

(3) Has the neuroscientist stated any credible theory of how human  memories could last for more than 50 years?

(4) Has the neuroscientist stated any credible theory of how humans could instantly retrieve a memory?

If none of these things have been done, then the chances are any claims that a neuroscientist has made insinuating evidence for neural memory storage are unfounded. 

An example of a paper that uses some of the tricks discussed above is the extremely misleading paper "Understanding the physical basis of memory: Molecular mechanisms of the engram." No actual evidence is given in this paper that any of the authors have any such thing as an understanding of a physical basis of memory or any robust evidence for a neural storage of memories. The authors main trick is to use over and over the never-justified word "engram," referring to a neural storage place of memories. No robust evidence has ever been produced for such a thing. The studies cited by the paper are mainly junk science studies with defects such as way-too-small study group sizes, a lack of a blinding protocol, and a lack of pre-registration, leaving the authors free to play "keep torturing the data until it confesses." 

Near the end of the paper the authors give some indications that they have no real understanding of a physical basis of a memory. We read these confessions:

"We are still far from identifying the 'double helix' of memory—if one even exists. We do not have a clear idea of how long-term, specific information may be stored in the brain, into separate engrams that can be reactivated when relevant. Understanding engram organization would be the equivalent to understanding how genes are organized in the genome....The long-term storage of memory information is still ultimately a mystery. Some theories point to storage at connectivity patterns between engram cells—controlled by processes, such as synaptogenesis, synapse elimination, or unsilencing of synapses, whereas others point to the storage of the information through diverse molecular mechanisms inside the cell....Despite the tremendous efforts in the field to clarify the molecular basis of memory processes, some very important questions remain unanswered:

What are the essential biochemical changes or features of engram cells that account for information specificity? How does one singular memory get recalled when it is behaviorally relevant?

How are the molecular mechanisms in memory systems not only flexible to allow immediate and quick learning but also persistent to store information during long-lasting periods? How are the systems compatible with allowing new learning involving modification of synapses while maintaining a trustworthy transcript of past experiences?

Where is the information stored for long-term? Is there such thing as a memory code? If the engram cells bear the mnemonic information, what is physically the substrate of that code? Can the experience of specific memories, and crucially not others, be transmitted as can be done with genetic information? And what is/are the associated mechanism/s that the code needs to function—where is the DNA double helix of memory information?

What is the role of traditionally less considered homeostatic processes, such as neurometabolism and thermodynamic efficiency in learning and memory?

Despite half a century of research, we are still only scratching the surface of the molecular basis for memory function."

With such confessions the authors give away they have no actual understanding of a physical basis of memory. 

The modern neuroscientist is rarely an impartial judge of truth. He or she is typically a vested interest, more like a juror who has been bribed to reach a particular conclusion. Living in a "publish or perish" university culture, the "life blood" of such neuroscientists is research funding. Such funding is obtained mainly from governments and corporations.  But there are no corporations nowadays with an impartial objective interest in the discovery of truth about minds and memories, but mainly instead corporations interested in developing particular products (typically devices or pills) based on conventional assumptions about brains. 

As for government funding, it is doled out by scientists themselves, who dole out research money in accordance with whatever belief dogmas prevail in particular scientist communities. There is no research money for heretical thinkers, but only research money for "follow the herd" conformists. To get the all-important research funding, the modern neuroscientist must kneel to the prevailing dogmas of his scientist community, and follow the defective old speech customs of his scientist tribe. When it comes to write up his paper, the neuroscientist's guiding principle is "make sure it conforms to current beliefs," and "make sure to claim some positive result," because papers reporting null results are unlikely to be published (something called publication bias). A person with such vested financial interests is unlikely to be an impartial judge of truth.  

Whenever a scientist attempts to persuade us that he understands something about how a brain could store memories, such as using little sound bites such as "synapse strengthening" or "changes in neural circuits," that scientist engages in misleading hand-waving.  Often neuroscientists will confess that they have no understanding of how a brain could store memories. Guillaume Thierry, a professor of cognitive neuroscience at Bangor University, recently stated the following about the brain:

"It’s not made up of data that can be transferred from one brain to another. It’s alive, and we don’t know why or how. If I wanted to transfer my memories into a machine, I would need to know what my memories are made of. But nobody knows.”

We can compare the modern neuroscientist (so often using electricity to try and show memory storage in brains) to a scientist named Bill trying to demonstrate that pigs can fly.  Imagine if Bill hooks up a pig to electrical wires, and repeatedly jolts the pig with electricity.  Most of the time he would merely get squeals of pain from the pig, but very rarely the pig might seem to rise up in the air a tiny bit, as all the voltage creates a little involuntary reflexive jumping effect. "That's it!" declares Bill triumphantly on his one hundredth attempt at jolting the pig with high voltage. "I've shown pigs can fly!" A scientist using a technique so goofy is like the modern neuroscientist who typically gets irrelevant and marginal results in his memory experiments, and then tries to pass that off as evidence of a brain storage of memories. 

No comments:

Post a Comment