You freeze in place, try to catch your breath, and in your head there is only one thought: "How did I do it?"
We have all experienced a similar situation. Although more often it is still about inadvertently turning on a new ultra-modern microwave oven, at random poking on the buttons. Whether you are saving your life or just want to reheat food, your brain needs to solve two problems at once in order to understand: action X entails result Y.
Artist problem: did I do it?
Action versus result problem: Which of the things I did caused the result Y?
The questions are not easy. We do a lot of things, and all this leads to something. In addition, some events are constantly happening around us, and only a small part of them depends on us. Therefore, the brain needs to separate the result Y from the general flow of events. Then he must determine if we have anything to do with what happened. At the same time, information from the senses comes only after performing actions that could have caused the incident. Dopamine, the first violin in symphonies of many cognitive theories, is responsible for these processes.
We have a hypothesis that describes in detail the neural process of correlating an action with its performer and result. This hypothesis comes from two fundamental ideas.
First, the brain has a model of how the outside world works - based on it, it constantly tries to guess what will happen next. If the forecast does not come true, then surprise arises, and the event that caused it stands out from the stream of ordinary and predictable phenomena.
Promotional video:
Secondly, the brain records everything that we have just done, which means that any unexpected event can be correlated with the chain of recent actions stored in memory. As soon as a connection is found, the action can be repeated - and check whether it will lead to a similar result. A positive answer will indicate a causal relationship.
In neither case can we do without our old friend - dopamine. At first glance, when it comes to correlating actions with results, this neurotransmitter is the worst of all possible helpers. Dopamine is produced in huge quantities in several areas of the brain at the same time. This method is completely ineffective to isolate a single connection between a set of neurons - say, between those responsible for the action X and the result Y. But in fact, this is an extremely sophisticated mechanism. The release of dopamine can be compared to broadcasting a radio signal. With its help, the following message is instantly sent to different parts of the brain: “Something very unusual happened just outside the Brain. How many of you guys will take responsibility for this?"
A person during this broadcast is surprised. This feeling occurs when the brain makes a mistake in its predictions. There is ample evidence that dopamine neurons serve to signal an error when the brain calculates the likelihood of receiving a reward. If your brain assumes that no reward will shine for you anytime soon, and suddenly a complete stranger hands you a donut, dopamine neurons are activated for a moment. They convey to the rest of the brain the surprise that something unexpectedly good has happened. The neurons seem to shout: "It doesn't matter which of you guys got us a donut, but it needs to be repeated!"
The brain can be wrong about more than just the likelihood of a reward. We also know that dopamine neurons are biased in predicting an unwanted outcome. Things you might want to learn to avoid, such as not pressing a button that triggers a snake dump into your bathroom. An incorrect assessment of the past tense after a recent event. And also that you are not singing quite the way you would like. You probably didn't know that you have a music critic sitting in your midbrain?
All of these mechanisms through which various errors trigger the short-term release of dopamine have a simple explanation: dopamine neurons are responsible for transmitting surprise. And, most importantly, this release always occurs immediately after an unexpected event Y and serves as its time stamp.
So, your brain has noticed that something cool has happened in the surrounding world, and dopamine notifies the rest of its parts about it. Now you need to determine if any of your actions were the cause of this turn. In this case, the brain, as it were, glues the action and the result, strengthening the local connection between them.
To do this, you need to find information about the action or actions that occurred before the information about the result was recorded. In the end, communication can only go from cause to effect, and not vice versa. Let's say a light comes on in the room - why? It’s unlikely because you marked the appearance of light with a special ritual dance on one leg and waving a dead chicken at the same time. Rather, the reason is that at the entrance you flipped the switch (of course, with the hand in which there was no chicken).
The main task of short-term dopamine release is to find the right one among the recent actions. When an electrical impulse begins to pass along the axon, carrying a message to recipient neurons, a long process begins inside the neuron, in which the concentrations of several molecules, in particular calcium, change. What's more, activity on any incoming connection to this neuron also leaves traces of calcium, marking this input as potentially important.
Dopamine also acts at the junction of two neurons. Suppose one neuron gave a command to perform an action that entailed a certain result, and another neuron, connecting with the first, reports: "I was activated at this time." Now the information is encoded in this connection: "Do the same when I am activated again." If the neuron responsible for the action is fired in response to the activation of the second neuron, then traces of calcium will remain in it. They will serve as a reminder that this particular connection and this particular neuron were involved. In the presence of calcium, the connection between these neurons will be enhanced by dopamine. Thus, the thought “do the same when I’m activated again” is only amplified if both neurons are activated at the right time.
Even more surprising is the fact that causality is built into the very rules by which the strength of the connections between two separate neurons changes. Apparently, the connection between neurons A and B remembers in which order they were fired. If neuron A is activated right in front of neuron B, then it could logically lead to the activation of the latter. This compound is labeled with calcium, and this bond can be strengthened in the future.
But if neuron A is activated immediately after neuron B, it can no longer be the cause of B activation. On the contrary, such a connection will need to be weakened, since in which case the activation of neuron A will interfere with neuron B. If neuron A is activated long before or long after neuron B, the strength of the connection will not change. Indeed, it seems that the rules for changing the strength of the connection are designed specifically to train the brain to establish causal connections.
This is how the brain solves the problem of correlating action with the result. He finds the action X that caused the result Y by broadcasting a signal that something unusual has happened outside the Brain and also by timestamping the event. This signal will only be received in the place where the neuron responsible for the action has just been activated. This is determined by the molecular traces that remain after activation. Now, if this connection fires again, the action neurons X are more likely to be activated. This means that the person himself in a similar situation is more likely to perform exactly the action X. This is how we determine whether X actually invokes Y, and tune our understanding of the external world.
It remains to solve the problem of correlating the action with the performer, and now this has become easier. How does the brain know that you have nothing to do with what is happening? The dopamine signal does not show any traces of activity in neurons. The absence of traces means: "I have nothing to do with it."
However, it can also happen like this: the neurons responsible for the action were activated immediately before the result, but were not its cause. This is why the action must be repeated. If action X is intentionally repeated and does not cause result Y, then there is no evidence that there is a connection between the two.
The principles by which the brain establishes causation is one of the main areas of work of modern neuroscience, but in general this area remains mysterious and little researched. Elements of the theory of perception of causal relationships from time to time surface in the literature, but the authors themselves do not focus on this. This means that in this area, hypothetically, it is possible to make many discoveries, given how many questions there are in it that have not yet been answered. Let's look at one of these questions. How does the brain use this information in the future?
The perception of causality is based on the idea that our brains use a predictive model of the world. If so, then we must also have an inverted model that answers the question "How to change the world?" We can say, “I want result Y,” and use the inverse model to find the required “action X” that will lead to the desired result.
This means that we need to constantly adapt two models: predictive (if you do this, this will change in the world) and inverted (for something in the world to change, you need to do this). It is highly likely that dopamine is responsible for tuning each of these circuits. But where does the adaptation itself take place? Do these models change together or separately? We have no idea about it. How many different models of the external world the brain creates, how they interact with each other and how they complement - these are all unanswered questions.
The ability to establish causal relationships through trial and error has been observed in different species. Not only in animals, but also in birds. This ability connects individual events in a sequence: if I do an action X, it will be followed by a result Y. Some species can establish causation through imitation. Observing their relatives, blue tits from the tit family can learn to unscrew the caps of milk bottles (seriously, it is better not to enrage these birds).
But man has one advantage - language. Thanks to him, we no longer need to waste energy on endless observations of chains of actions, limited only by our own experience. With the help of language, we can explain causal relationships and convey them in the abstract: in books, magazines, documentaries. Or take a multi-hour YouTube guide on how to go over a V8. We can record our observations by leaving spaces where there are not enough links in the chain between X and Y (this is called "science"). We can share information and find causal relationships on a larger scale and in larger samples than is available to an individual.
The fact that humans have identified the causes of complex phenomena such as species extinction or global warming is evidence of our ability to comprehend the world beyond individual experience. Only the human brain is able to understand not only what caused itself, but also what we all caused.
Mark Humphries