Chasing One’s Shadow
Reflecting on the existence of the universe, the origins of life, and why life became conscious is something most people do now and then—some more than others, and perhaps especially when young. This is likely why some—mistakenly, in my view—consider Fyodor Dostoevsky’s books to be “for the young,” when life is more about finding one’s place than filling it with things. Even if our minds are perhaps more dynamic in youth, we should strive to maintain an open attitude and curiosity about life even later on. For me, this is what it means to be a researcher.
The question of the universe’s origin was what led me to study physics at university, and the mystery of consciousness what later turned me toward neuroscience and medicine. (The origin of life, for some reason, was less interesting to me.) Working clinically with patients suffering from severe somatic and psychiatric illnesses instills humility. Morning rounds filled with suffering—parents who never wake up in intensive care, or young adults steadily losing themselves to delusions and passivity. The fact that we have any means of alleviating this is nothing short of remarkable, given how little we truly know about the brain.
In patients with neurological or organic disorders, unconsciousness is not expected unless large parts of the cerebrum or brainstem are damaged. At the back of the brainstem lies the reticular formation, a network controlling arousal and other vital systems. It is also home to small clusters of cells producing monoamines (e.g., dopamine), the receptors and transporters of which are targeted by most psychotropic drugs. If one suffers a bleed or inflammation in these areas, consciousness may diminish—along with more local symptoms depending on which tracts pass through or originate there.
In psychiatric patients, however, there is often nothing visible. One may be deeply psychotic—unable to form coherent sentences or maintain hygiene—yet show no abnormalities on the most advanced brain imaging. Isn't that a mystery? Something is clearly wrong with brain function—with how billions of neurons and synapses cooperate. But where does it begin? Is the whole brain affected? Is there a difference between being a little psychotic and severely so?
This ties into the question of when and where consciousness arises. In the famous Libet experiment, researchers tried to measure when a person becomes aware of their decisions.
Libet claimed to show that unconscious brain activity preceded the conscious decision to move. In that way, someone monitoring brain activity could predict the subject's will before the subject themselves was aware of it! Can this really be true? Later studies have claimed that brain activity predicting a voluntary act can be detected up to 10 seconds before the action.
The late philosopher Daniel Dennett argued that the interpretation of Libet’s experiment is flawed. Beyond methodological issues (such as determining exactly when a person becomes aware of a decision), Libet’s conclusions are conceptually wrong. The reasoning assumes we can locate the precise start of a cognitive process. But just as a wave is inherently dynamic, so is volition. It's incorrect to imagine thoughts becoming conscious at a fixed moment and playing out on a stage—what Dennett called the Cartesian theater, after Descartes’s dualism—where our “inner self” watches. The problem then becomes: who exactly is that inner self? And so begins an infinite regress.
Trying to find the beginning of a thought is like chasing your own shadow. It’s impossible, and eventually you get tired and feel a bit silly.
Free will consists of both conscious and unconscious processes. What matters is that the individual feels agency and has veto power over impulses that do not align with their will. This is the core of the argument that we can live in a deterministic world at the macroscopic level and still have free will. We need not resort to hand-waving about quantum mechanics or chaos.
From an evolutionary perspective, it must have been advantageous to develop an increasingly complex brain that—at some phylogenetic point—became conscious. The individual, as carrier of the species' DNA, could better navigate threats and seek mates and food. Personally, I find it hard to believe that consciousness is exclusive to Homo sapiens—or that it’s even possible to draw a line where one species becomes conscious and another is not.
To be conscious, one must be able to distinguish between what has happened to oneself and to others, and recall personal past events. This is called episodic memory, as opposed to semantic memory. I remember where and when my children were born—not just as facts. Likewise, a conscious being can imagine the future, simulate scenarios, and plan ahead.
Swedish research has shown that not only great apes like chimpanzees and orangutans, but even ravens, can plan for the future! However, it’s not certain that ravens truly plan in the human sense; some researchers argue it's merely associative learning, where the birds learn during training that certain behaviors lead to favorable outcomes. This is analogous to artificial intelligence, where deep neural networks associate input with outcomes through reinforcement learning.
The Brain and the Self
Free will, then, is a sense of agency—that I, not someone else, want something. I’ve written previously about how primary psychotic disorders like schizophrenia are tied to disturbances in self at a fundamental level. That said, one cannot claim that people with schizophrenia lack consciousness. Of course they are conscious—but perhaps it is qualitatively altered? How this happens is still an unanswered question.
Attempts to understand the brain as an organ that simulates the external world and compares predictions to perception have been made by, among others, Karl Friston, professor at University College London. His Free Energy Principle treats "free energy" not as a physical concept, but an information-theoretic one.
We can imagine the brain having a model of the world, m, which it compares with actual sensory input, s. The surprise (greater when input is unlikely given the model)—equivalent to Shannon information—can be written: −log p(s∣m).
We assume that the sensory input s arises from a latent variable x. The probability of receiving s given model m is, by Bayes’ theorem:
p(x∣s,m)=p(x∣m)*p(s∣x,m) / p(s∣m)
However, this probability distribution cannot be directly estimated. Instead, we minimize free energy, which serves as an upper bound on surprise. That is the essence of the Free Energy Principle.
Put simply: the brain builds a model of the world and constantly updates it to minimize the discrepancy between predicted and actual input. This has interesting connections to psychosis, where the brain may be too liberal in its interpretations of the world (I often tell patients that their brain’s filter is too leaky). This is called salience—the weight we assign to sensory input. This is why the phrase "psychosis as a state of aberrant salience" has been coined. Dopamine signaling (in specific synapses) can be viewed as regulating how much importance we assign to different inputs.
For me, it's natural to think of the brain as a prediction machine, embedded in an environment to be understood and anticipated. Yet this says nothing about the subjective experience of agency or the feel of a qualia—like the redness of a rose—as perceived by me. Perhaps this is more a philosophical issue, but no less valuable to contemplate—regardless of age.
Well written, Erik!
Enjoyed this piece a lot Erik!