The Biggest Brain Projects Are Still Failing to Support Prevailing Brain Dogmas

In a December 2020 post I examined the failure of the two biggest brain research projects to back up claims commonly made about the brain, such as the claim that the brain produces the mind and the claim that brains store memories. Let us now look at how one of those two biggest brain research programs (the Human Brain Project) is still failing to substantiate such claims. The Human Brain Project is a billion-dollar European research project. 

The page here of the Human Brain Project web site is entitled "Highlights and Achievements," and presumably lists the biggest accomplishments of the Human Brain Project. Let's take a look at the items listed at the top of the page, in the year 2021 section.  The first five items merely discuss technology innovations, not anything involving new findings about the brain.  The sixth item is merely an interview with a professor who talks about no specific research findings of the Human Brain Project, and who says that the project has "become a truly enabling endeavor," which is the kind of vague praise that people give when they don't have much in the way of specific achievements to discuss.  Then we have an item merely talking about how humans have some brain cell types not found in mice.  

The next item is entitled "Controlling brain states with a ray of light." We have a statement of never-substantiated neuroscientist dogma:  "The brain presents different states depending on the communication between billions of neurons, and this network is the basis of all our perceptions, memories, and behaviors."  But the page discussing this ray of light research mentions nothing that sounds important.  We merely hear of some light being sent into a brain, with some transition occurring, although the only transition claimed is an awakening from sleep: "This new chemically-engineered tool allowed to induce and investigate in detail, in a controlled and non-invasive way, the transitions of brain from sleep- to awake-like states using direct illumination." Not very impressive, given that we have already long known of a tool for inducing a transition from sleep to awake-like states: the humble alarm clock.

The next item merely mentions work on some robot.  The item after that has the title "EBRAINS powers brain simulations to give insight into consciousness and its disorders." The page discussing this research mentions no progress in understanding how consciousness occurs. It merely mentions some project reading brain waves during normal consciousness and sleep. We have a quote making it sound as if unconsciousness always involves less complex brain waves:

"We can see that unconsciousness is not simply a matter of a loss of brain activity,” Massimini says. “It’s not necessarily weaker. But it is a lot less complex.” 

This statement is only half-true. Brain waves are less complex for patients under anesthesia. But the most complex brain waves are those seen during grand mal seizures (also called tonic-clonic seizures), and during such seizures people are typically unconscious. An EEG reading during a grand mal seizures resembles a seismograph reading during an earthquake. 

The next item is entitled "HBP-researchers find new approach for Energy-Efficient AI Applications," which obviously involves no progress in cognitive neuroscience.  The item after that merely involves brain surgery, not cognitive neuroscience.  The next item merely is something pertaining to spinal cord surgery. 

We then see an item of little significance, merely something about some new technique for modeling dendrites. The item after that is the claim "A new means of neuronal communication discovered in the human brain." The claim is unjustified, being based solely on a paper failing to prevent robust evidence. 

The paper is the paper "Long-range phase synchronization of high-frequency oscillations in human cortex."  The claim of a synchronization effect is not well established.  The paper looked for correlations after analyzing brain wave readings from fewer than 100 people.  A paper like this would only be credible if (a) it was a pre-registered study that declared before any data was gathered a hypothesis to be tested, how the data would be gathered and how the data would be analyzed, and (b) the paper discussed a thorough blinding protocol that was followed.  But there is no mention of any pre-registration of this study, and the paper never mentions any blinding protocol (failing to use the word "blind" in its text).  

So what was going on? Apparently the authors got some EEG readings, and were then absolutely free to analyze the data in any way they wanted, being free to slice and dice the data until they found something they could call "synchronization."  We should have very little confidence in a study following such a method.  Given a body of data and freedom to analyze it any of 1001 ways, it is all too easy to find "synchronization" that is no real effect. For example, if I can compare the wins and losses of sports teams with the ups and downs of stock markets, options markets and bond markets, I could probably find  a little something I could claim as "synchronization." 

While the Human Brain Project site has bragged that "a new means of neuronal communication" has been discovered, the scientific paper behind this claim does not even sound very confident of such a thing, merely saying that some brain oscillations "may be synchronized between widely distributed brain regions." Also, neuron communication does not mean that neurons make our minds or store our memories. 

The last item on the Human Brain Project's list of 2021 highlights is merely a discussion of some paper claiming similarities in the brains of birds and mammals.  We read a claim that "the brains of birds and mammals look surprisingly similar in their organization." This is not at all true, and bird brains look very different from human brains. 

Judging from the Human Brain Project's list of 2021 highlights, the lavishly funded Human Brain Project is not making any progress in verifying the main dogmas of cognitive neuroscientists, the claim that the brain is the source of the human mind, and the claim that brains store memories.  Similarly, we find no support for such dogmas in a recent article entitled "The Human Brain Project: six achievements of Europe’s largest neuroscience programme."

Here are the six achievments listed:

• "Human brain atlas":  We read about merely fancy descriptions of parts of the brain. 
• "Synapses in the hippocampus:" We read that "researchers have published detailed 3D-maps of around 25,000 synapses – electrical and chemical signals between brain cells – in the human hippocampus." Such a result does not seem so impressive when you consider that the brain is believed to contain trillions of synapses. Also, you don't explain mental phenomena such as understanding and memory by making maps of synapses or maps of neurons. 
• "Robot hands":  Obviously this has nothing to do with verifying the claims of cognitive neuroscientists.
• "A neuro-inspired computer":  The computer described is not anything like a computer having the characteristics of the brain. If you ever built such a computer, it would never work to process data reliably and at high speeds. In digital computers electrical signals travel with 100% reliability, but in the cortex of the brain a signal will only pass across a synapse with a likelihood of 50% or less. Computers have coordinate systems and indexing systems allowing the computer to instantly find the location of some stored data, but brains have no such things. 
• "Virtual epileptic patient":  This has nothing to do with verifying the claims of cognitive neuroscientists.
• "Scientific output":  We merely hear a mention that 1497 papers cite the Human Brain Project. 

In the year 2020 section of the "Highlights and Achievements" page of the Human Brain Project, you won't find anything that substantiates the main dogmas about brains taught by neuroscientists. My December 2020 post here discusses the items in that section (as well as the 2019, 2018 and 2017 sections), and explains why they fail to support claims such as the claim that brain make minds and the claim that brains store memories. 

The Human Brain Project is making no progress in supporting claims such as the claim that brains make minds and the claim that brains store memories because such claims are not correct.  But what about the other big brain project, the US-based BRAIN Initiative? In my December 2020 post I examined the failure of that project (as well as the Human Brain Project) to back up claims commonly made about the brain, such as the claim that the brain produces the mind and the claim that brains store memories.  Were there any big results for the BRAIN Initiative in 2021?

Apparently not, judging from the page here which lists 2021 highlights for the BRAIN Initiative.  There is some discussion of brain mapping that has not yet done anything to back up the main dogmas of neuroscience. We see only two stories relevant to whether brains make minds:

• A story entitled "Neuroprothesis restores words to man with paralysis."
• A story entitled "Reading Minds with Ultrasound: A Less-Invasive Technique to Decode the Brain's Intentions."

The first story discusses some man who had a stroke leading to brain stem damage causing him to lose the power of speech. Electrodes were planted in his head, to look for some correlation between motor cortex brain activity and attempts of the man to say one of 50 different words. A system was developed wherein the man's attempts to speak can be matched to one of the 50 words.  This merely shows that the brain has a role in the muscle movements related to speech.  It does not prove that the ideas for what to say arise from the brain. 

The story about "reading minds with ultrasound" has a title that is misleading clickbait. The corresponding study was merely done with monkeys.  What's going on is some obscure clear-as-mud business involving trying to predict which of two options (left or right) a monkey will take, based on reading brain states a few seconds before the movement. A good rule of thumb for experimental science is to ignore all studies that did not use at least 15 subjects per study group.  The main results for this study involve experiments on only a single monkey. The study (which shows no sign of using a blinding protocol) is not reliable evidence for any ability to read minds with ultrasound. 

It appears that neither the Human Brain Project in Europe nor the BRAIN Initiative in the US is making progress in supporting claims such as the claim that brains make minds and the claim that brains store memories.  Such progress will never be made because the brain is not the source of our mind, and our brains do not store memories. To find reasons justifying these statements, read the other posts on this blog. 

In today's science news, we have the results of a project to test the reproducibility of cancer research.  A paper reports little success in reproducing results.  We hear that a large fraction of scientists simply refused to respond to queries from fellow scientists trying to reproduce the results, which is just what we would expect if a significant fraction of published research was fraudulent or defective. Here is a very worrying quote from the abstract:

"We conducted the Reproducibility Project: Cancer Biology to investigate the replicability of preclinical research in cancer biology....However, the various barriers and challenges we encountered while designing and conducting the experiments meant that we were only able to repeat 50 experiments from 23 papers. Here we report these barriers and challenges. First, many original papers failed to report key descriptive and inferential statistics: the data needed to compute effect sizes and conduct power analyses was publicly accessible for just 4 of 193 experiments. Moreover, despite contacting the authors of the original papers, we were unable to obtain these data for 68% of the experiments. Second, none of the 193 experiments were described in sufficient detail in the original paper to enable us to design protocols to repeat the experiments, so we had to seek clarifications from the original authors. While authors were extremely or very helpful for 41% of experiments, they were minimally helpful for 9% of experiments, and not at all helpful (or did not respond to us) for 32% of experiments."

Can you imagine a more damning statistic about the work quality of today's biological researchers, the fact that "none of the 193 experiments were described in sufficient detail in the original paper to enable us to design protocols to repeat the experiments"?

In a separate paper, the researchers found that "the median effect size in the replications was 85% smaller than the median effect size in the original experiments, and 92% of replication effect sizes were smaller than the original," which suggests a high degree of unreliability in biomedical research.