How quick is a human thought

Thoughts are free - but will they stay that way?

I keep thinking. I have tried many times to stop my thinking; to create a void in me in order to then learn more about the structure of my thoughts by comparing this void with my everyday thinking. I never succeeded. In meditation, it is said, this can be achieved after long practice. One day I will probably have to take on the training of meditation in order to get to the bottom of my thinking in an inner emptiness.

I observe in myself that my thinking is constantly changing its character. Sometimes it is blurred beyond recognition; a dull chaos of splintered thoughts and wordless images that line up and overlay. Sometimes, for brief moments, I am just seeing, feeling, or hearing. Sometimes my mind suddenly jumps to something new and I don't know why. And at times my thinking is crystal clear. A lucid thought then carries me through the complex web of arguments for hours and I easily recognize the internal structure of the object on which I am working mentally. At such times, thinking is an intoxicating pleasure. As a psychologist and brain researcher, I try to understand the neural basis of thinking. For this research we need the whole methodological range of the cognitive neurosciences. In cell cultures, for example, hybrid compositions of nerve cells and microchips are tinkered with in order to conduct a currently still primitive biological-technical dialogue with small groups of nerve cells. In animal experiments, a wide variety of animal species learn to solve tasks devised by researchers, while at the same time one records the activity of dozen of their nerve cells and tries to decipher the subtasks that the individual neurons carry out.

These experiments reveal that neurons function like small cogwheels in a huge machine by taking on sub-tasks of a large functional structure. In other experiments, scientists reconstruct the complex processes in the human brain and manage to isolate individual building blocks of thinking and their associated neural signatures. In clinical studies, paralyzed people are equipped with electrodes in or on their brains to enable them to control wheelchairs and robotic arms solely through the power of their thoughts or to communicate with their surroundings through writing. All these findings help us to understand better and better how thinking, learning, remembering, making decisions and acting work and why these processes sometimes fail. Basic research, driven by curiosity, makes the later clinical application possible. Once we have deciphered the psychological and neural signatures of thinking, we can help many people with illnesses and disabilities. But can we then also read minds? Could all these findings possibly lead to the touching folk song “The thoughts are free” only being sung with a bitter aftertaste because our thinking has become transparent?

Nothing is as important to research as a theory that guides experimentation and helps the researcher convert data obtained into real knowledge. Probably the most fundamental theory of cognitive neuroscience was formulated in 1949 by the Canadian psychologist Donald Hebb in his book "The Organization of Behavior: A Neuropsychological Theory". Hebb specifies three postulates that still serve as the basic pattern of today's neuroscientific research.

The first postulate states that neurons that are active together (and thus “fire” together in the jargon of neuroscience) develop more effective synapses with one another. I want to explain this with an example. Let's imagine that you have moved to another apartment and are cooking there for the first time in your new kitchen. While the pan sizzles, lean forward quickly to grab a spice. Some of the nerve cells in your brain are currently processing the situation: “I'm standing in front of the stove”, “Spices are on the shelf in front of me”, “I'm reaching for the spice jars” and so on. At that moment you collide painfully with the extractor hood. Other nerve cells report immediately: "Pain on the forehead", "Extractor hood hangs lower than in the old kitchen" and so on. All of the nerve cells listed in this fictional scene now fire together for a brief moment. This makes the synaptic bond between them stronger. A stronger synaptic bond means that the next time you cook on your new stove, the nerve cells that process their current situation (for example, “I'm standing in front of the stove”) are active again. However, the activation of these neurons is now able to activate those nerve cells that processed the painful collision at the time due to the strong synaptic contacts. This allows you to remember how painful it used to be while you were cooking and that you need a new movement pattern in order to season your pan without wounds.

Donald Hebb's first postulate (neurons that fire together, wire together) has proven to be absolutely correct neurobiologically. As simple as this postulate sounds, the proposed solution to a fundamental problem in brain research is just as ingenious: how does the brain organize itself and how does it integrate the experiences made in life without the existence of a superordinate control system that tells the brain how to do it should? Today we know that according to Hebb’s rule, synapses are strengthened by correlating the activity of simultaneously firing neurons. As a result, the formation of memory in our brain is organized through the joint occurrence of events, which are then neuronally associated.

For you, the reader of this article, that means that I am in the process of changing your brain. Millions of neurons in your nervous system are currently processing the content of this page. The synapses in which both nerve cells involved are successfully active at the same time therefore go through a complex chain of molecular processes at the end of which the strengthening of these synapses stands. If you remember this article tomorrow, I have successfully modified your brain.

Hebb’s second postulate is that nerve cells form into flexible, short-term firing coalitions (so-called assemblies) that then represent an object, an intention to act or a thought. It is important at this point to define exactly what is meant by a neural coalition. For example, a Neuron A can be part of the “stove” assembly, also fire a few minutes later in the “desk” assembly and shortly afterwards be silent when you think about your car. On the other hand, a neuron B could possibly remain inactive with “stove”, but fire with “desk” and “car”. However, should you learn something new about your desk, the compositions can change, so that, for example, Neuron A ceases to be a member of this assembly. If you have never heard the term assembly in this context, a new constellation of nerve cells may form in your cerebral cortex, which through their joint activity increases the synaptic efficiency within this group (first Hebbian postulate) and associations with it established other similar terms (that is, with assemblies that had sprung up in your brain earlier).

Every time you hear or read the word assembly in the future, or if you think about the neural correlates of thinking, you will activate this very new constellation of nerve cells. And if you come to new insights in this reflection, your assembly for the term “assembly” will also change in the composition of its neural members.

It is important to note that the neurons that make up an assembly do not necessarily have to be spatially adjacent. On the contrary, they are likely to be spread across different areas of your cerebral cortex. Take the assembly for the stove in your kitchen. “Herd” is a word in the German language and so a number of neurons in the language area of ​​the left half of the brain will be part of the “Herd” assembly.

However, your focus also has a certain appearance and therefore nerve cells in the area of ​​your visual system will participate in this assembly. Since you often operate the buttons on your hearth, nerve cells near the motor centers of your hands will also be part of the “hearth” assembly. Donald Hebb was probably largely right with his second postulate, although definitive evidence for Hebb’s assemblies is still pending. Even if the concept of assemblies is still controversial at the moment, neuroscientists agree that large groups of neurons are active in changing combinations when thinking. These activity patterns migrate rapidly across the surface of the cerebral cortex, with the same thinking content usually being associated with similar activity patterns. This enables neuroscientists to understand what a person is thinking about to a certain extent.

But since every brain is much more unique than a fingerprint, a computer first has to learn the activity patterns of a particular person's brain. To do this, the person is placed in a scanner and the experimenter shows them either an A or a B on a monitor several times. Each of these letters leads to a specific activation pattern in the brain, which is learned by a computer. Now the experimental set-up can be changed: the test subject is still sometimes shown an A or a B, but the experimenter no longer knows which letter is currently appearing on the monitor. He now has to guess this from the brain's activation patterns.

Mind reading works quite well at this simple level. One can take these experiments so far that one can extremely roughly understand what a person is thinking about as they begin to dream, or which of two alternatives they will decide on in a few seconds. For the success of all these studies, however, the test subject must always be confronted with a set of stimuli beforehand so that the computer can learn the individual activation patterns of each person's brain for each stimulus. Hebb’s third postulate states that assemblies are arranged in sequences in such a way that the end of the activity of one assembly marks the beginning of the activity of the next. This could possibly represent the neural basis for the uninterrupted stream of thoughts that we all experience. Verifying the correctness of this postulate is a difficult task. In fact, it has been observed that nerve cells in areas such as the hippocampus, which is important for memory formation, are organized in temporally arranged stages and could function like clocks for assemblies in other areas of the brain. And there are many studies that show that neurons are active in small circuits in repetitive sequence patterns. The problem with current brain research is not so much the small repetitive circuits. The question is rather which mechanism assemblies use to organize themselves into novel sequences in a flexible way. Most likely, many assemblies are competing to be next in the chain. How the selection of the next assembly succeeds and how the constant overlapping of different assembly chains can be prevented is part of the puzzles that have not yet been satisfactorily resolved.

Exploring the neural foundations of thinking is probably the most fundamental challenge in neuroscience. Our ability to think complexly has made us human, and thinking disorders are a central feature of many brain disorders. A great deal of basic scientific research is still necessary in order to understand the neural principles of thinking to the extent that the core causal problems of the various diseases of the brain can be clarified.

Until then, most therapeutic procedures in neurology and psychiatry will have to alleviate the symptoms instead of eliminating the causes of the disease. But the neuroscientific research into thinking has, in addition to many findings for clinical application, also created a by-product that can mean a dramatic increase in the quality of their lives for many patients. People with complete or extensive paralysis are currently dependent on their environment for their care and for the fulfillment of the simplest wishes. As explained above, however, signatures of intentions to act can be identified in every brain. You don't even need to use a large scanner, but can use simple electrodes that are stuck to the scalp to derive the neural correlates of intentions to act. By systematically training pattern detectors, technical systems will later be able to maneuver wheelchairs, for example. For more complex actions, such as those performed by robotic hands, small electrodes must be implanted either on or in the patient's cortex. This gives patients a technical third hand with which they can do many everyday things.

If we are already able today to read simple pictures, words and decisions from the brain of test subjects, do we have to fear that we will soon become spiritually transparent? Where are the limits of the technical and scientific development of mind reading? Should the resolution of scanners or electrophysiological methods improve in the near future, the quality of the neural signal read will of course also increase. Current scanners with very high magnetic field strengths are, however, already close to the physically sensible upper edge of the resolution. The electrophysiological methods will very likely never achieve this resolution. In other words, we have not yet reached the limit, but we are approaching the technical limits that these technologies bring with them. With the constantly increasing computing capacity of computers, it could possibly be possible within the next one to two decades, not only broad categories of thinking (“man”, “street”, “car”), but also more differentiated thoughts, such as those of one capture a certain scene, a specific person or a word. Complex trains of thought could still not be grasped. In addition, one must mention a boundary condition in these considerations.

All previous examinations only work with highly cooperative test subjects who look at stimulus material for hours without moving so that the scanner can learn their corresponding brain activities. Later, in the actual test phase, these test subjects adhere to the test protocol and, for example, think precisely of the picture or word that the computer is supposed to capture in their brain. As soon as these systems are to be used to convict potential criminals, it will probably become clear which mental counter-strategies can be developed by people who do not want to reveal their thinking.

But what if tomorrow a technological revolution gave us a completely new tool with which we could record the activity of practically all neurons? Such a fictitious scenario can hardly be answered sensibly. But I suspect that even such a high resolution system cannot solve the problem of complete mind reading. This problem lies in the correlation between the mental and neural signals. In such a scenario, the theoretically infinite number of mental processes of a person are also opposed to an almost infinite number of combinations of neuronal signals. These must first be mapped to one another. To do this, the test person would have to think an extremely large number of different thoughts and communicate them precisely so that the pattern detectors learn the associated neural signals. How long does it take until you have thought so many different things and told what you thought, before the pattern detector can read out of me what I want to keep to myself? And then there is another big problem, which was outlined at the very beginning: My thinking is nowhere near so clear that I can always communicate it precisely. I am only aware of part of my thinking and I can only express part of my conscious thinking in words. The rest of my thinking is inaccessible even to me, but it contributes to the neural signals that future systems could capture.

I think it stays that way: the thoughts are free.

© All rights reserved. Frankfurter Allgemeine Zeitung GmbH, Frankfurt. Provided by the Frankfurter Allgemeine Archiv