Friday, June 1, 2018

When the King of Science Soundbites Misses the Mark

Anyone who frequently watches science-related TV shows will instantly recognize the face of physicist Michio Kaku, who very frequently appears in such shows, giving slick little soundbites. Having a mellifluous speaking voice, Kaku must be rated A+ as a TV talking head. He's the king of the science soundbites. But are his science books as good as his TV soundbites?

Kaku's book The Future of the Mind is not one that I can rank highly, because I think it is built on a fallacious premise: the idea that the mind is merely the product of the brain. Kaku engages in a wide variety of speculations of how humans might be able to improve their minds by using things such as neural implants. In a similar vein, he speculates about uploading minds into computers or beaming minds into interstellar space. These speculations are fun to read, but very dubious because they are all based on the very dubious idea that the mind is purely a by-product of the brain.

Kaku makes no serious attempt to establish this idea; he just takes it as a given. A serious attempt to establish this idea would be a massive undertaking, as it would need to tell us things such as these:

  • how immaterial consciousness could arise from material neurons;
  • how a brain could possibly generate a thought or idea;
  • how brains could store memories for 50 years despite the fact that proteins in synapses have lifetimes of only a few days or weeks;
  • how complex memories and complex ideas could be stored in neurons, things which show  no physical sign of encoding stored symbolic information (other than genetic information in DNA);
  • how miscellaneous neuron firings could possibly create such a thing as a unified self;
  • how humans could instantly recall obscure items of information given a brain with no coordinate system and no position notation system (presumably needed to instantly find an exact location where some minor memory was stored).

Such an attempt would also need to explain away the very many lines of evidence for psychic phenomena suggesting the mind is more than a brain product. Kaku doesn't attempt to do any of this. What Kaku does repeatedly in this book is to make dubious citations of neuroscience research. I'll give some examples.

On page 14 to 15 Kaku tries with remarkable chutzpah to cite the case of Phineas Gage as something that caused a “paradigm shift in scientific thinking,” discrediting the idea of the soul. Gage's case certainly did no such thing, and citing it as evidence against such an idea makes no sense at all. What happened to Gage is that he accidentally got a thick railroad spike (1.25 inches in diameter) driven through the front of his brain. But a doctor examining him later declared that he was "quite recovered in his faculties of body and mind," and he went on to live a pretty normal life.  This case (along with countless other cases of high mental function despite huge brain damage discussed here and here) actually help to discredit the idea that minds are merely the product of brains, helping to shore up the likelihood of something like a soul. Some who knew Gage reported he underwent a personality change, but there are many reasons why personalities may change, and we don't know whether his personality change was related to his injury. An expert on Gage has stated that the personality change “did not last much longer than about two to three years,” but Gage lived for 12 years after the accident.

On page 16 Kaku claims this about the neuroscientist Penfield: “Penfield found that by stimulating parts of the temporal lobe, his patients suddenly relived long-forgotten memories in a crystal clear fashion.” It's actually not known whether visual images arising from such stimulation are memories or hallucinations. A review of 80 years of experiments on electrical stimulation of the brain uses the word “reminiscences” for accounts that may or may not be memory retrievals. The review tells us, “This remains a rare phenomenon with from 0.3% to 0.59% EBS [electrical brain stimulation] inducing reminiscences.” The review states the following:

We observed a surprisingly large variety of reminiscences covering all aspects of declarative memory. However, most were poorly detailed and only a few were episodic. This result does not support theories of a highly stable and detailed memory, as initially postulated, and still widely believed as true by the general public....Overall, only one patient reported what appeared to be a clearly detailed episodic memory for which he spontaneously specified that he had never thought about it....Overall, these results do not support Penfield's idea of a highly stable memory that can be replayed randomly by EBS. Hence, results of EBS should not, at this stage, be taken as evidence for long-term episodic memories that can sometimes be retrieved.

This does not at all match what Kaku has suggested, there being only one example in 80 years of the type of thing he has suggested (recall of long-forgotten memory through electrical stimulation). I may note that no such cases prove that memories are stored in brains. If you rub an ice cube on my foot, I may remember some time in my youth when my feet once felt cold; but that doesn't mean memories are stored in my feet.

On page 34 Kaku tells us, “The concept of 'I' as a single unified whole making all decisions continuously, is an illusion created by our own subsconscious minds.” This very dubious claim is not empirically established,  and Kaku certainly does not validate it by citing a silly claim by Steven Pinker that consciousness is like a storm raging in the brain.

On page 37 to 39 Kaku attempts to cite split brain research to support his assumption of the mind being the product of the brain. As discussed here, the actual research (correctly interpreted) does not lend support to such a claim, but contradicts it. When the fibers connecting the hemispheres of the brain are severed, we are left with a single mind, not two (although such single minds have been inaccurately described as “two minds” by certain people eager to support the “minds come from brains” dogma).

On page 43 Kaku offers a strange definition of consciousness: “Consciousness is the process of creating a model of the world using multiple feedback loops in various parameters (e.g. in temperature, space, time, and in relation to others) , in order to accomplish a goal (e.g. find mates, food, shelter).” That's way too complicated a definition. You don't need to be creating a model of the world or trying to accomplish a goal to be conscious. A better definition of consciousness would be this: "the mental facility needed for a mind to have a life experience consisting of things such as sensory perceptions, thoughts, or feelings.”

Scientists lack any coherent explanation for how a brain could generate thought or intellect. Scientists often fall back on localization claims to try to hide this shortfall. A scientist who cannot explain the how of a brain making an intellect will often try to use a where as a substitute, by suggesting that specific mental capabilities come from particular parts of the brain. So on page 56 Kaku tells us that the dorsolateral prefrontal cortex is “the CEO of the brain.” There is no good evidence for such an assumption. A scientific paper tested the IQ of six patients with lesions in the dorsolateral prefrontal cortex, and found that they had an average IQ of 104.8, above the average of 100, exactly the opposite of what we would expect if such a region was the “CEO of the brain.” See this post for a wide variety of neuroscience evidence conflicting with the claim that intelligence or decision making comes from the prefrontal cortex.

On page 64 Kaku makes an important misstatement about ESP. He states, “True telepathy, found in science fiction and fantasy novels, is not possible without outside assistance.” This is untrue. As reported here, decades of experimental research have found extremely dramatic evidence for extrasensory perception, without outside assistance. Such experimental evidence (replicated many times) is inconsistent with claims that minds are merely the products of brains. We would expect to have zero evidence for telepathy if such claims were correct. The mainstream peer-reviewed science journal American Psychologist just published a review of the evidence for psi (phenomena such as ESP), a review that stated this:

The evidence provides cumulative support for the reality of psi, which cannot be readily explained away by the quality of the studies, fraud, selective reporting, experimental or analytical incompetence, or other frequent criticisms. The evidence for psi is comparable to that for established phenomena in psychology and other disciplines.

Scientists do not understand how a brain could store memories. But you would never know this from reading pages 106 to 107 of Kaku's book. He states as if it were fact a completely speculative and extraordinarily unbelievable theory about memory storage in brains: a kind of “splice and scatter” theory. His claim is that memories to be stored first go to the hippocampus, and then are spliced up and distributed in different parts of the brain – kind of like what would happen if you put a page through a shredder, and then scattered the paper strips in different parts of your house. Here is how Kaku states this very unbelievable idea:

To store these memories for a longer duration, the information must then run through the hippocampus, where memories are broken down into different categories. Rather than storing all memories in one area of the brain like a tape recorder or hard drive, the hippocampus redirects the fragments to various cortices....For instance, emotional memories are stored in the amygdala, but words are recorded in the temporal lobe. Meanwhile, colors and other visual information are collected in the occipital lobe, and the sense of touch and movement reside in the parietal lobe.

Such claims are not maintained by most neuroscientists, and there is no good evidence to support them. Tests have shown that patients with damage to the medial temporal lobe have “fully intact” remote autobiographic memories – not what we would see if words related to memories are stored in the temporal lobe. We still have no convincing evidence that any autobiographical or episodic memories are stored in the human brain. When we look at a recent review by neuroscientists of the current thinking about a brain storage of memories, we find near the end of the paper a very conspicuous uncertainty, with the scientists shooting out speculations in all different directions, as if they had nothing like firm ideas on how memory could be brain-stored – something quite contrary to the impression Kaku has given us that this matter has been figured out. Similarly, in this recent review of the current thinking about a brain storage of memories, we are told “our current knowledge of how such memories are stored in the brain and retrieved, as well as the dynamics of the circuits involved, remains scarce.” Using the term “engrams” to mean the hypothesis that there are cells in the brain that store memories, the authors state “what and where engrams are implicated in remote memory storage and how they change over time have received little experimental attention thus far.” The authors also frankly tell us that “ From engrams to spines surprisingly little evidence exists in the literature on the grounds of remote information processing, maintenance and storage to account for the lifelong and persistent nature of the mnemonic signal.”

Let me explain why the “splice and scatter” theory Kaku advanced is untenable. Let us first consider a simpler idea: that all the information needed to recall some particular memory is stored in one specific location of the brain – for example, your memory of your first kiss may be stored at neuron number 345,212,232. This then creates the extremely grave problem I call the navigation problem, the problem discussed here that there would seem to be no way for the brain to navigate instantly to that exact location – because the brain would not know where that location was. Finding that exact spot in a brain with no position notation system and no coordinate system would be like trying to find an exact page in a book in a library of a million volumes, none of which had titles on their covers. Or, to give another analogy, it would be like trying to find just the right post office box in a vast post office with millions of boxes, none of them numbered. But we are able to instantly recall obscure memories from decades ago. Someone may name some obscure figure you learned about many years ago, and you can instantly recall some facts about that person. No neuroscientist can explain how this could happen. The navigation problem is one of three main reasons for thinking that our episodic and conceptual memories are not actually stored in our brains, the other two being the permanence problem discussed here (that there is no viable theory of how memories could be stored for decades in a brain with such high protein turnover), and the encoding problem discussed here (which is that no one has a decent theory of how episodic or conceptual memories could ever be translated into neuron states).

memory retrieval


But let us suppose, ala Kaku, that a single memory is not stored in only one spot, but is somehow spliced up and scattered among multiple spots. This does not make the navigation problem any less severe; instead, it makes it much worse. For now, rather than having to instantly find just one exact spot (one needle in the haystack), the brain would have to instantly find multiple needles in the mountain-sized haystack. This would be like instantly retrieving the information in seven books scattered around a million-volume library in which none of the shelves were marked and none of the books had titles on their covers.

In addition to such problems, Kaku's speculation creates the problem that we cannot conceive of any way in which a memory could be spliced up and its parts copied to different regions, nor can we conceive of any way in which scattered fragments could be reassembled into a memory that we can instantly recall as a unified whole. In a system in which tiny parts have location numbers, you can splice up information and scatter it. For example, I could scatter a 10-page note across ten post office boxes at a post office, by having a system whereby the first page in one post office box ended with a remark such as “continued in post office box 879,” and the second page in box 879 ended with a remark like “continued in post office box 1293.” But such a thing couldn't be going on in a brain, where none of the neurons have neuron numbers, each neuron is connected to about 10,000 other neurons, and there is no coordinate system, numbering system, or position notation system. Given limitations in the brain such as the lack of a unique pathway between one neuron and another neuron and the complete absence of any coordinate system, position notation system or numbering system, it is impossible to imagine any way that a brain could ever instantly recall a memory that was spliced up and stored in scattered regions of the brain.

Repeatedly in this book, Kaku proves to be an uncritical consumer of neuroscience hype. An example is on page 108, where he talks about an experiment with mice. He states the following:

Scientists at Wake Forest University and the University of Southern California made history when they were able to record a memory made by mice and store it digitally in a computer. This was a proof-of-principle experiment, in which they showed that the dream of downloading memories into a brain might one day become reality.

The actual experiment was something very minor. The scientists recorded some brain signals that were being passed about in a mouse's head while it did a particular simple task (flipping a switch). When such signals were replayed to other mice, they were somewhat more likely to remember something. This is kind of perceptual playback, involves short-term memory only, and tells us nothing about whether long-term memory is stored in human brains. The experiment was not at all recording a memory in the sense of reading a memory permanently stored in the brain.

Kaku's most recent book is entitled “The Future of Humanity.” It's a pretty good book, and I only object to occasional passages in it. One such passage is on page 270, where Kaku says, “At present all known physical phenomenon, from the Big Bang to the motion of subatomic particles, can be explained by two theories: Einstein's general theory of relativity and the quantum theory.” This untrue statement is triumphalist nonsense. For one thing, it implies that Einstein's general theory of relativity explains the Big Bang, the universe's beginning; and that theory certainly does not at all explain why there was such an event. For another thing, there are countless physical phenomena that are not explained by either quantum theory or general relativity, including the origin of life, the origin of species, and earthquakes.

In fact, general relativity doesn't even well explain the speeds at which stars orbit the center of a galaxy, and scientists try to patch up this shortfall by bringing in a whole other theory, the theory of dark matter. As a news story states, For decades, astronomers have noticed that the behavior of galaxies and galaxy clusters doesn’t seem to fit the predictions of general relativity.” The modern scientist needs at least ten theories to try to explain physical reality, and still he does not do a very good job of explaining such a reality.

Another misstatement occurs on page 279 where Kaku states, “Supersymmetry is the largest symmetry ever found in physics,” making it sound as if physicists have actually discovered supersymmetry in nature. They certainly have not. To the contrary, supersymmetry is a floundering theory that has done nothing but fail empirical tests.

In the book Kaku discusses string theory, his unproven physics specialty which thus far has not proven to be either useful or empirically successful. But don't worry, Kaku has an idea on how it may be a little bit useful. On page 287 he says, “The hope is that using string theory, we will be able to calculate the quantum corrections needed to analyze exotic phenomena such as time travel, interdimensional travel, wormholes, and what happened before the Big Bang.” This is kind of like claiming that your “white elephant” brain wave feedback gizmo will soon be able to cure depression, double your sexual pleasure, triple your intelligence and give you immortality.

Offering a completely groundless and gigantically extravagant speculation as if it were fact, Kaku also states on page 305, “The explanation for why the universe seems to be fine-tuned to allow for life as we know it is because of luck, because there are billions of parallel universes that are not fine-tuned for life, that are completely lifeless.” But evoking such other universes (which is an act of pure fantasy) is pointless. If you want to believe that our universe is fantastically lucky, you can do so without assuming such other universes; so why bother to assume them? The multiverse is not an explanation. An explanation is something that makes a particular thing certain or likely. If we assume a trillion other universes, this does not increase the chance of our universe being life-compatible. And why is Kaku is telling us matter-of-factly that “there are billions of parallel universes” when there is no evidence at all for even one such universe other than our own?

It would seem that the “king of science soundbites” Kaku does not always get things right. And I doubt that opinion would change if I were to read Kaku's book “Parallel Worlds,” on the “fake science” topic of parallel universes. But on this page I can find some quotes from that book, one of which is the following:

At first, it seems as if the existence of complex life forms on Earth violates the second law. It seems remarkable that out of the chaos of the early Earth emerged an incredible diversity of intricate life forms, even harboring intelligence and consciousness, lowering the amount of entropy. Some have taken this miracle to imply the hand of a benevolent creator. But remember that life is driven by the natural laws of evolution, and that total entropy still increases, because additional energy fueling life is constantly being added by the Sun. If we include the Sun and Earth, then the total entropy still increases.

There are no "natural laws of evolution" in the sense of anything like the law of gravitation, and when scientists refer to natural selection they refer merely to a statistical effect, that fit organisms reproduce more often than unfit organisms. The origin of life was not at all driven by any law of evolution, since natural selection cannot occur until life originated. The argument about the sun by Kaku is what mathematician Granville Sewell calls the "compensation argument." It's the idea that you can have events seemingly defying the second law of thermodynamics in one place, as long as they are "compensated" by high-entropy events elsewhere. In this paper he forcefully argues that the argument is fallacious. He states the following:

Of course the whole idea of compensation, whether by distant or nearby events, makes no sense logically: an extremely improbable event is not rendered less improbable simply by the occurrence of ‘‘compensating’’ events elsewhere. According to this reasoning, the second law does not prevent scrap metal from reorganizing itself into a computer in one room, as long as two computers in the next room are rusting into scrap metal — and the door is open.

No comments:

Post a Comment