By Matthew Cobb March 11, 2021 ⋅ 24 min read ⋅ Books
Introduction
Steno’s view: if we want to understand what the brain does and how it does it, we should view the brain as a machine and take it apart to see how it works.
For over 350 years, we’ve been following this suggestion.
This books tells the story of the brain and the discoveries that we’ve made along the way.
Despite a solid bedrock of understanding, we have no clear idea on how neurons produce the brain’s activity.
Throughout history and even now, the brain has been explained by analogy to that era’s most advance technology.
E.g. Clock, hydraulic machine, telegraph, and now computer.
The brain isn’t a computer, but it’s more like a computer than it’s like a clock.
Over the centuries, each layer of technological metaphor has added something new to our understanding.
But by holding tightly to metaphors, we end up limiting what and how we can think about the brain.
E.g. If brains are computers, then we ignore the fact that it’s a product of evolution.
Metaphors shape our ideas in ways that aren’t always helpful.
This book isn’t a history of neuroscience, but a history of the brain.
Two problems with understanding the brain
The brain is mind-bogglingly complicated.
Even though we have vast amounts of brain data, we’re in a crisis of ideas about what to do with all that data; about what it all means.
One of the most tragic indicators of our underlying uncertainty about the brain is the real crisis in our understanding of mental health.
Part I: Past
The history of science is rather different from other kinds of history, because science is progressive.
Every idea, no matter how outdated, was once modern, exciting, and new.
We should never dismiss past ideas because one day we will become the past too.
We are simply doing the best we can, just as our forebears did.
Chapter 1: Heart
For most of history, we have viewed the heart (and not the brain), as the fundamental organ of thought and feeling.
E.g. We used to attribute emotions and feelings to the heart using words like heartbroken and heartfelt.
This has applied across most cultures, not just Western culture.
This is because heart-centered views match our everyday experience; the heart changes its rhythm at the same time as our feelings change.
E.g. Anger, love, and fear all cause the heart to beat faster.
People believed this idea because it made sense.
The first recorded challenge to our heart-centered view occurred in ancient Greece at around 600 and 250 BCE.
However, Aristotle dismissed the idea as there was no actual proof (at the time) of any link between thought and the brain.
To Aristotle, this makes logical sense.
The acceptance and permission to directly study human anatomy enabled us to make significant discoveries with regard to the brain and the nervous system, as the dissection of the human body was previously forbidden.
The heart-centered view shows us how everyday experience and intuition can mislead us from the truth.
One researcher, Galen, became certain that the brain was the center of thought and proved it with his experiments on pigs.
E.g. Galen would expose the pig brain and when pressure was applied to a certain area, the pig would stop moving.
For over 1200 years, ventricular localization was widely accepted. It’s the idea that mental functions are located in the brain’s three ventricle (open spaces).
E.g. Sensation and imagination in the front ventricle, intellect and reason in the middle, and motion and memory in the back.
Faith, not fact, was still the essence of knowledge and formed the framework of European intellectual life for a good part of history.
Vesalius could understand the body because careful dissection can reveal structures such as bone, muscle, and blood vessel, which imply different functions.
However, dissection of the brain leads to no obvious difference in its structure as it looks mostly uniform.
Vesalius couldn’t find any qualitative difference between the structure of a human brain and that of other vertebrates.
There was no single ‘brain-centric moment’ when thinkers realized that the brain, and not the heart, was the key organ.
Chapter 2: Forces
The shift in attitude was slow and complex; there was no single experiment or dissection that pushed the argument towards the brain.
Review of Descartes’s ideas about the brain and mind, most of which are false.
Swammerdam’s experiments showed that neither the pneumatic nor the hydraulic models of nerve function were correct.
The organization of the brain was presumed to reflect its function.
Steno’s approach to the brain, of taking it apart and attempting to identify the functions of those parts, is more or less what we’ve been doing ever since.
Thomas Hobbes, Margaret Cavendish, and Princess Elizabeth all agree that Descartes’ view of the mind as immaterial was wrong, and that the mind is based in matter.
One argument against materialism was that since all matter is made of atoms, then the atoms involved in thinking matter must have some special quality; but all atoms must be fundamentally identical, so the stuff that makes up the brain can’t be special in any way.
The flaw in this argument is the assumption that thinking matter must be special; it isn’t.
Unless we mean the special matter to be neurons, but the argument only applies to atoms, not living cells.
While philosophers worried away at the metaphysics of the mind, physicians and other investigators addressed the apparently simpler question of how perception and movement occurred.
Review of the reflex.
La Mettrie’s realization that “the ability to think is merely the consequence of the organization of the machine”.
George Prochaska claimed “As the spark is latent in the steel or flint, and is not elicited, unless there be friction between the flint and steel, so the nervous force is latent, nor excites action of the nervous system until excited by an applied stimulus”.
As we now know in hindsight, this claim is referring to the action potential.
This conditional, non-mechanical view raised the question of how any known force could fulfill such a role.
Exciting hints about what this latent force might be came from a new phenomenon that seemed to be linked with life itself: electricity.
Chapter 3: Electricity
When a dispute can’t be resolved, new evidence is needed to make progress.
Review of Galvani’s experiments on frog legs.
Aldini’s experiments suggested that electricity was more than just an irritant, but that it was the source of complex behavior.
Nerve function bore a greater resemblance to the transmission of electricity along a conducting wire than to any other fact that we’re familiar with.
Two problems with the slow speed of an action potential
The brain can only respond to events in the past. However, this isn’t a big issue as our sense organs are close to the brain. Otherwise, our consciousness would significantly lag behind the present.
An explanation as to why the speed of electrical activity in nerves was far slower than that in wires.
We live, just slightly, in the past and never in the present.
By the middle of the nineteenth century, the assumption that brain structure was related to brain function had become deeply rooted in popular imagination.
Chapter 4: Function
Review of phrenology.
Searching for biological principles by comparing different species has turned out to be a very powerful method in science.
The brain as the organ of the mind.
The mind appeared to be distributed across the cortex.
Review of Broca’s and Wernicke’s work.
Ferrier reported that he failed to observe any responses from electrically stimulating the frontal regions of the brain in the monkey, cat, and dog.
This fitted with the account of Phineas Gage who had a metal pipe destroy part of his frontal lobe.
While there was nothing obviously different, Phineas’s personality had changed.
Chapter 5: Evolution
Review of Darwin’s theory of evolution.
Nobody could even begin to describe how consciousness might emerge from the activity of the brain.
Every principle gyrus and sulcus of a chimpanzee’s brain is clearly represented in that of a man.
As Darwin explained, the behavioral and intellectual differences that existed between primate species must, in some way, reside in slight anatomical differences, and not in some large structure present in one group and completely absent in another.
The brain of an ant is one of the most marvelous atoms of matter in the world, perhaps more so than the brain of a man.
Darwin had revealed a principle that could explain why the brains of different animals are different shapes, because they had evolved to produce different behaviors.
One argument for Materialism is that even though we may not currently understand a phenomenon, that doesn’t mean that we’ll never be able to understand it.
To argue that there are things that we can never understand is to undermine the whole point of science, which is to explain the unexplained.
As the wave of doubt on materialism swept through Europe, the key insights about the brain from Darwin were forgotten.
Chapter 6: Inhibition
It’s known that nerves can make things happen, like muscle contraction, but it can also stop things from happening.
This was made apparent in the middle of the 19th century.
E.g. Stimulation of the vagus nerve decreases heart rate and can even make it stop.
This demonstrates that the brain can have an inhibitory action on movements.
Further evidence comes from the destruction of the cerebral hemispheres, which leads to unrestrained reflex action.
This makes clear that the brain contains the cause for the hindrance in the activation of the nervous principles.
Despite all of the interest in inhibition, it wasn’t clear how it actually worked (neurotransmitters weren’t known at the time).
One of the overall functions of the brain is to control the body.
Both Freud and Pavlov, while great figures, their ideas had no consequences for how we understand the brain.
Perceptions aren’t simple impressions produced by the environment, but rather inductive conclusions, unconsciously formed.
E.g. Our impression of a 3D world is constructed by our brains out of two 2D images without us being aware.
Weber and Fechner showed us that our ability to perceive the differences between two stimuli changes with their amplitude.
E.g. The heavier two objects are, the larger the difference between them has to be before we can detect any differences.
Another way of putting it is that we’re very good at detecting small differences between low-amplitude stimuli.
Review of our blind spot in visual perception.
The implication was that the complex structures of the brain were somehow able to perform logical operations not only without conscious thought being involved, but apparently as a prerequisite to that conscious thought.
Helmholtz’s view of the brain as an active organ and that perception as imperfect was a major breakthrough in our understanding of what the brain does.
For the basis of brain activity to actually be understood, scientists first had to realize what the brain was actually made of.
Chapter 7: Neurons
One of the greatest scientific achievements of the nineteenth century was cell theory; the realization that all organisms are made up of cells, and that cells can only come from other cells.
This marked a major step in progress as the brain was now known to be made up of cells, specifically nerve cells called neurons.
Camilo Golgi discovered silver nitrate staining by accident when he spilled some of it onto brain tissue hardened with potassium dichromate.
The tissue looked black to the naked eye but under the microscope, only a small portion of nerve cells had been stained.
Since only a few cells were stained, this provided support that nerve cells were actually cells and not some continuous network like in some other organisms.
Review of Cajal’s work and the history behind the neuron doctrine.
The telegraph metaphor was prominent in the nineteenth century.
The principle that nervous current could only go in one direction was obvious when it came to the microscopic organization of the sensory systems such as the retina.
Another case where nervous current only goes in one direction is in reflex arcs.
E.g. Tapping the tendon below your knee causes your thigh muscle to contract, but you can’t stimulate your thigh to make your tendon respond.
We see that mental exercise may accentuate transmission along certain, more specific, routes in accordance with skills that have been learned.
At the beginning of the 20th century, it was clear that some kind of electrical charge was being sent down neurons, but it was less clear what happened next.
Cajal had no proof of what happened when two neurons met, nor how the current was transmitted.
Every synapse isn’t a passive gap, but instead offers an opportunity for a change in character of nervous impulses.
The impulse from the axon terminal differs from the nerve impulse in the dendrite.
It was hypothesized that perhaps the surfaces of the axon and dendrite might hold the secret to what happens when a nerve impulse moves across them.
The explanation of synaptic function apparently lay in the structure of the membranes of the two cells involved.
The water-and-valve analogy used to explain electricity was also used to understand nerve impulses as action potentials weren’t known at the time.
However, this analogy broke down at the synapse.
Review of the disagreement between the electrical (sparks) camp and chemical (soups) camps of synaptic transmission.
In hindsight, both are true, but the soups camp won out at the time.
Review of neurotransmitters.
Biological discovery was outstripping the dominant technological metaphor and revealing that the brain isn’t a telephone exchange.
Other metaphors were going to be necessary to understand what the brain does and how it does it.
Chapter 8: Machines
The eighteenth century suggestion that humans were machines was flipped during the twentieth century as it seemed that machines would become human.
Review of behaviorism.
At the time, there was still no good explanation of apparent goal-directed phenomenon in physiology and behaviour.
Uexkull’s two key ideas
Umwelt: the inner sensory world of each species.
That the brain could sense how its output had changed the world and alter its functioning accordingly; in essence feedback.
The nature of our sensory impressions is determined a priori, before any experience, by this physiological apparatus of our senses, sensory nerves, and sensory nerve centers.
One way to be relatively sure of understanding a mechanism is to make that mechanism.
At the same time as these models were being built out of wires and metal, neurophysiologists were realizing that real nervous systems worked in a very different way.
As had been expected, the electrochemical transmission of a nerve impulse was very different from the movement of electricity down a telegraph cable or a telephone wire.
Biology was proving to be more complicated than technology.
Review of the discovery of the refractory period and the all-or-none response.
Three major discoveries by Adrian and Zotterman
Sensory neurons respond in an all-or-none manner like muscles.
If a neuron is repeatedly stimulated, the cell will stop responding.
When a neuron fires, the amplitude and shape of the response, soon known as spike, is constant, but the frequency of the firing changes with stimulus intensity.
Adrian’s research on nerve function provided evidence for a clear correlation between the activity of neurons and perception.
As with most scientific discoveries, if Adrian hadn’t done this work, someone else would have done so at about the same time. That’s the nature of science.
The sensory messages were scarcely more complex than a succession of dots in Morse code.
This new way of thinking about what neurons do pointed to a whole new area of research on how to understand and break the neural code.
This is also when nerve messages were thought to contain information.
To Adrian, the whole point of a nervous system was to transmit encoded information about the world along neurons.
The significance of Adrian’s realization, that there was a neural code and that this message contained some kind of information, were part of a transformation of our understanding of how nervous systems and brains work.
Chapter 9: Control
Review of McCulloch and Pitts (MP) work.
In the 1930s, the use of the words ‘feedback’, ‘circuit’, ‘input’, and ‘output’ were becoming more common to describe biological systems such as the brain.
The collaboration between MP lead to the common metaphor now used to explain the brain: as a computer.
Except things didn’t happen this way, as the link was actually the other way around by thinking that a computer is a brain.
The key insight from McCulloch was that the all-or-none response from a neuron resembled the logical true or false.
This made it possible to understand the activity of neurons in terms of a series of logical propositions.
By intelligently organizing neurons, we could make AND, OR, and NOT logic gates out of neurons.
Sensory illusions demonstrate the clear dependence of the correspondence between perception and the external world upon the specific structural properties of the intervening nervous net.
The real novelty of MP’s work was that it focused attention on processes rather than on anatomical regions.
The key idea was the relation between the component parts and the way that function emerged from organization.
However, one major issue with these models was mapping them to real biology; a model so far removed from reality seemed pointless.
The simplified MP neurons didn’t exactly match what was found in real neurons.
Key ideas from Craik
Technology was effectively an extension of human sense-organs and bodies.
The brain aims to build a mental model of its environment so that it can prepare the organism for future events.
In 1946, von Neumann became skeptical of the parallels between computers and brains as the brain was more complicated than thought.
E.g. Firing rate increases with stimulus strength, which doesn’t fit the computer metaphor.
When it comes to representing the outside world, neurons aren’t digital.
While no significant insights about the brain were gained post-World War II, there was significant consensus on the idea that the activity of the human brain and the existence of the mind as the same thing.
During the first half of the twentieth century, we explored what the brain does, but not how it might do it. That was left to the second half of the twentieth century and brings us to the present.
Part II: Present
No major conceptual innovation has been made in our overall understanding of how the brain works for over half a century.
We have not made progress on the view established by Craik and McCulloch that the brain contains symbolic representations of the outside world that it manipulates to predict what will happen next.
The term ‘neuroscience’ first appeared in the 1960s and gained widespread use in the 1970s.
There are now tens of thousands of neuroscientists around the world in a variety of fields such as cognitive, theoretical, and clinical neuroscience.
However, this part will skip over research on sleep, non-visual perception, hormones, emotions, development, and genes.
Paradoxically, despite the immense progress, it isn’t clear if we have the theoretical tools necessary to face the challenges of understanding the brain in the twenty-first century.
Chapter 10: Memory
Review of Wilder Penfield’s experiments on stimulating the temporal lobe.
These experiments suggest that memory might have a very precise location in the brain.
The experience of a memory is played back through the same networks that recorded it.
Somewhere deep in the brain, linked to very particular parts of the temporal lobe, memories could be evoked by electrical stimulation of a very precise area.
However, these evoked recollections were very different from ordinary memories as they had far more detail and had a dream-like quality.
One thing was clear: there wasn’t anything special about the particular evoked memories.
Recent studies have confirmed the accuracy of Penfield’s experimental work.
The area Penfield stimulated wasn’t where memories were stored, but instead the area seemed to be able to trigger the activity of distant parts of the brain where the memory was actually stored.
This complicated the debate between localized and distributed function.
Other primates show a different sensory and motor homunculus, suggesting that it’s a consequence of our evolution and biology.
The homunculus is misleading because it’s the average response of all patients, but any person might show a different relation (aka a different homunculus).
Review of Hebb’s work and patient HM.
The fine structure of the nervous system, the network of connections between cells, is formed through experience.
The destruction of the hippocampi results in the inability to form new memories.
The hippocampi don’t store memories but rather are required to create them.
Review of Tolman’s rat maze experiments.
Somehow, the brain represents the outside world in its neurons.
Review of O’Keefe’s place cell discovery.
His findings suggest that the hippocampus provides the rest of the brain with a spatial reference map.
So the hippocampus encodes episodic memories, places, and how to get from one location to another by simulating movement and the subsequent location.
It’s a map, but a cognitive one that involves multiple sensory modalities and is based on associations and predictions.
It isn’t a simple one-to-one representation of the outside world.
Also, these maps vary between species depending on their niche and sensory organs.
Review of Britt and Moser’s discovery of grid cells.
All of these intriguing results show that the hippocampus integrates information of various kinds with different objectives.
At the same time, other brains regions are involved in the creation and recall of a particular memory, suggesting a mixture of localization and distributed function.
Different regions of the brain are involved in creating our memories.
It may be that whenever we remember something, we’re using the memory palace technique.
An interesting finding that supports this idea was that HM couldn’t compare two odors and he found it difficult to read a map.
In humans, olfactory perception and spatial memory are fundamentally intertwined in the hippocampus.
Review of Eric Kandel’s work and Hodgkin and Huxley’s work.
Kandel proved that Hebb’s neurophysiological postulate was correct; that learning involves a change in the synaptic strength in small circuits of neurons.
Short-term memory involved enhanced release of a neurotransmitter, while long-term memory involved the growth of new synaptic connections.
The engram was nothing more than a change in the activity of a synapse.
Review of long-term potentiation (LTP).
As long-term memory is created, new synaptic connections are established through the creation of new dendritic spines, rather than existing spines changing their shape.
Shrinking the spines disrupted memory, indicating that dendritic spines are a key component of engram formation.
However, it’s becoming apparent that neurons don’t create new synapses on their own and rely on astrocytes to promote synaptic plasticity.
If the activation of astrocytes is blocked, then memory is impaired.
Chapter 11: Circuits
Review of Hubel and Wiesel’s (HW) findings.
In the visual cortex, each column corresponded to an object (line, dot) and each layer corresponded to a particular orientation of that object (vertical, horizontal).
The same goes for the somatosensory cortex with different columns for different body parts, and different layers for different kinds of sensory stimuli from the same body part.
Review of the critical period and how the brain changes structure as a consequence of experience.
HW’s work suggests that visual processing in the brain is hierarchically organized.
Review of grandmother cells, the Jennifer Aniston neuron, and the dorsal and ventral visual streams.
Sparse coding: the higher the level of representation, the fewer the number of cells involved and the sparser their activity.
Review of the connectome and our inability to have cell-level connectome knowledge due to technological issues with the storage and processing of brain scans at that resolution.
Work by researchers in simple nervous systems suggests that connectomics, while providing essential anatomical background, will fail to explain what’s going on in the brain unless it’s also accompanied by both experimental and modelling approaches.
The complexity of any nervous system is simply astonishing.
There’s a need to develop a theory of how neural circuits operate, but like our predecessors, it’s unclear as to what the next steps forward are.
We don’t have an appropriate framework for understanding the brain.
Part of the problem in analyzing how brains work is the complexity that can be found in the behavior of the simplest neural circuits.
The tight link between circuit structure and a particular output turns out to not exist.
E.g. Marder’s group showed that there were many different sets of activity in individual neurons that could produce the same overall pattern when they were connected together.
You can’t assume that the same behavior involves the same structure or pattern of neuronal activity as functions can be multiply realized.
E.g. Worms don’t all behave in exactly the same way, unlike a set of machines with the exact same wiring diagram.
Although simply observing components and their relations will not explain how a system works, describing the nature of the relations between those components and how they affect each other does provide a basis for explaining how the system functions.
David Marr realized that Barlow’s five dogmas were missing something underneath, that something was wrong.
Marr argued that the best way to understand the brain was to copy its key abilities.
Chapter 12: Computers
Perhaps some aspects of nervous system circuits simply emerge from very basic rules.
Review of the Pandemonium model, perception, connectionism, and Marr’s three levels of analysis.
Simple cells don’t detect edges but construct them.
Evidence indicates that there are no Jennifer Aniston cells or individual cell detectors.
But there is an area that’s involved in recognizing faces in the temporal lobe.
Remember, there is no mini-monkey in the brain peering at the output of neurons. Somehow, the whole system produces perception.
Review of parallel distribution processing (PDP) and deep learning (DL).
Although machine, monkey, and human were all able to identify pictures of animals, the computer made very different errors from those made by the animals, suggesting that it wasn’t processing the images in the same way.
Despite these examples of machine learning, these computer programs don’t produce clear biological hypotheses and as a result, shed little light on how real brains work.
The brain wasn’t designed and as a result, we can’t be sure that it’ll embody ‘deep general principles’.
Review of the disappointment of the Human Brain Project, and Spaun by Chris Eliasmith.
Maybe we don’t need to model every neuron to understand what’s happening in the brain.
Over the last two decades, many neuroscientists have become increasingly convinced that the brain works along the lines of Bayesian logic.
Review of Karl Friston’s free energy principle.
There’s experimental evidence that our perceptions can be subject to top-down influences that’s required by Friston’s model and by Bayesian approaches in general.
The gap between theory and neurobiological evidence of the precise activity of single cells can be seen in the study of apparently simple predictive systems.
E.g. When insects find a mate, they need to intercept them while they’re flying.
The existence of something like Bayesian prediction in the brain seems certain.
However, experimental evidence will always be key to determine the validity of any theory, no matter how elegant or seductive it might be.
Electrical stimulation of the brain can produce or stop movement.
The lesson is that scientific research doesn’t take place in a vacuum and that exciting discoveries and therapeutic opportunities may have profound and unforeseen consequences.
Chapter 13: Chemistry
Review of Hofmann’s discovery of LSD and the massive shift towards experimenting with drugs and substances.
Their discovery marked the beginning of a profound shift in approaches to mental illness; away from the psychoanalytic and towards medicine.
These drugs have given scientists new ways of understanding the chemistry of brain function.
It still isn’t known how lithium exerts its very real effects on patients with severe mania.
All of these new drugs had two aspects
Clinically significant
Promise of a radical new insight into how the brain might work
These results reinforced the idea that the brain isn’t only an electrical machine but also a gland.
New concepts were needed to explain the chemical complexity of the brain.
The word “neurotransmitter” wasn’t coined until 1961.
The discovery of GABA resolved the problem of how inhibition occurs in neurons.
Two classes of neurotransmitter receptors: fast and slow.
Neurohormones are involved in the long-term control of essential physiological processes, many of which have a behavioural component.
Dopamine is involved in some addictions but not all.
It isn’t clear that the activation of dopaminergic neurons even produces a pleasurable sensation.
In most cases, the causes of mental health problems are hard to explain in terms of brain function or chemistry.
One popular attempt has been the link between serotonin and depression.
However, we don’t know the mechanisms behind this link and if there’s even a link at all.
It seems unlikely that there’s a single explanation and a single best treatment for depression, or any other mental health problem.
The other explanation for mental health problems, genetics, has little relevance for understanding the brain too.
We don’t understand how a healthy brain and mind work, so it’s hardly surprising that we don’t know how to fix things when problems arise.
Ultimately, it doesn’t matter if we don’t understand how and why a treatment works as long as it does work.
Chapter 14: Localisation
A repeated theme in our long quest to understand the brain has been the claim that certain functions are localized to certain brain structures.
Review of the invention of the CT scan, PET scan, and fMRI.
The brain in an fMRI scan isn’t a computer nor a neural network but a gland.
The fMRI revolution had begun.
fMRI activity is indeed tightly linked to neuronal activity.
fMRI reports a simple measure of the brain’s physiology but those images don’t directly describe anything like the actual activity of the brain’s neurons.
fMRI is unable to reveal one of the key aspects of how the brain works: the difference between activation and inhibition.
The fundamental weakness of fMRI is that it’s too coarse to allow for a real understanding of the computational activity of the brain.
Despite its immense ingenuity and astonishing technical prowess, fMRI studies haven’t significantly contributed to our understanding of how the brain works in terms of creating a model of its overall activity.
The next time you read a claim that a particular ability, emotion, or concept has been localized to a particular region of the human brain using fMRI, ask yourself, “So what?”
There’s a deeper issue at stake with approaches that seek to localize functions to particular structures.
Lesion studies and neuroanatomy clearly show structural differences, but the problem lies in the logic behind linking structure to function.
Just because a particular cell is activated doesn’t mean that’s all the cell does.
Because of the complexity of nervous systems, causal-mechanistic explanations are qualitatively different from understanding how a combination of component modules performing the computation at a lower level produces emergent behavior at a higher level.
Claims of localization of function are generally overextrapolated. At best, we have identified a location necessary for function and have often merely shown a correlation between location and function.
Even the direct manipulation of a particular cell or network alters or restores a given function, but that still doesn’t mean that the function is located in that structure.
What it does show is that the structure is required for the function, which generally involves a large network of neurons.
E.g. The amygdala is often said where fear is located but it’s more nuanced than that.
Localization of function has become blurred and more complex than originally thought.
However, one clear area for localization is in sensory processing.
E.g. There’s no evidence that olfactory signals are processed in the primate visual cortex and vice versa.
Brain function involves both segregation and integration.
One false but popular brain idea is that the brain is divided into a reptilian brain and human brain.
Another false idea is that of mirror neurons.
Different structures can apparently give rise to the same function.
Chapter 15: Consciousness
Review of Francis Crick’s work and the neural correlates of consciousness.
The corpus callosum enabled the transfer of learning of all kinds between the two hemispheres.
Review of split-brain patients.
Splitting the brain does seem to split the mind.
Except for language and emotion, there are no clear fundamental differences in function of the two sides of the brain.
The brain doesn’t work as two separate halves but as an integrated whole.
The enigma is that even though the cerebellum has more neurons, we don’t consider it to be involved in the processes of consciousness.
No one can yet explain why the activity of one set of neurons produces consciousness, but another set of neurons doesn’t.
One significant advance in our understanding of consciousness was that there are unconscious processes in the brain.
Review of David Chalmers easy and hard problems of consciousness and Thomas Nagel’s “what it is like to be a bat” paper.
As Crick warned, philosophers play by different rules from scientists.
These philosophical views are really a confession of despair.
Not one piece of experimental evidence directly points to a non-material explanation of mind.
In the last decade or so, the insights provided by Hebb and Crick into how to study consciousness scientifically by focusing on precise, solvable problems seems to have been somewhat forgotten.
Although there are many ways of theorizing consciousness, there are currently two main scientific approaches, the global workspace theory and integrated information theory, neither of which is widely accepted.
Quantum approaches to understanding biological phenomena are attractive to some people, partly because of the assumption that if two things are mysterious, then they may be linked, but there’s no evidence that quantum mechanics can explain consciousness.
Clarity will come when we have a more solid basis to build upon.
It’s unlikely that there will be a single experiment or theory that demonstrates how brain activity becomes conscious.
Just like how our thinking shifted from the site of thought in the heart to the brain; there was no brain-centric moment then, it seems unlikely that there will be a similar shift for consciousness in the future.
Instead, the slow accumulation of evidence will gradually shed light.
Part III: Future
Neuroscience still largely lacks organizing principles or a theoretical framework for converting brain data into fundamental knowledge and understanding.
Our understanding of the brain appears to be approaching an impasse.
What’s the next metaphor to be used to understand the brain?
We should dismiss with any metaphors regarding the brain and focus on it alone.
In the past, we’ve used new kinds of technology. This implies that the appearance of new and insightful metaphors for the brain hinges on future technological breakthroughs.
However, there is no sign of such development.
Review of weak and strong emergence as a way to explain the brain.
Attempts to reverse-engineer the brain are doomed to failure, as one researcher argued because the starting point is almost certainly wrong: there is no overall logic.
The inability of neuroscientists to understand a microprocessor given our current approaches hints that we still need to make significant theoretical breakthroughs.
One of the greatest problems that we have to solve is how to link the cell-level of the brain with the population-level of the brain.
The author suspects that for much of the rest of the century, this is the main problem we face.
An even bigger problem with theories that are intended to understand brain function on the basis of structure can be seen if we imagine that a microprocessor were an example of Crick’s alien technology.
Crick’s alien technology: if we were given a device from a Martian/alien, how could we understand what it does and how it does it?
A full analysis would reveal that input from the external world could alter its function, but it seems unlikely that we would realize that a Martian would use the device to play a game.
And to know that, we would have to see how a Martian interacts with the machine.
In the absence of that external element, both the meaning and the mode of functioning of the device would be obscure.
The author argues for developing analytical techniques and theoretical frameworks for understanding smaller and simpler systems before applying those lessons to humans.
Small brains also enable us to investigate how the structure of the brain is a function of two kinds of history: developmental and evolutionary.
Evolution has solved all of the problems we’re interested in, we just have to find those organisms and figure out how to ask them how they did it.
As we understand more, localization of function will become increasingly blurred and imprecise, and brains will be understood primarily in terms of circuits and their interaction, rather than on the basis of anatomical regions viewed as modules.