Let me start with a disclaimer: we don’t know how the brain works exactly and we probably won’t know in the foreseeable future. We don’t know how the brain goes from neuronal activity to the diversity of human behavior. The methods that we have to study the brain provide tons of data but a unified theory of the brain doesn’t seem to be near.
But we do know some things about the brain.
We know how neurons and synapses work. We know about the brain’s functional and structural hierarchical organization. We know that it does parallel processing. We know that it uses very little energy to do all it does.
And we know it resembles a computer in many ways. But, is it a computer?
We’ll understand the brain by looking at computers
The metaphor of the brain as a computer has been discussed by neuroscientists and computer scientists for decades. And its controversy still holds today.
Some argue that the brain is, in fact, a computer. They usually subscribe to this idea not because of its inherent truth, but because of its usefulness. A computer takes inputs, processes them, and provides some outputs. The same does the brain. Thus, there may be some ideas we could take from computer science and use them to provide knowledge about the biological machine the brain is.
Gary Marcus, professor of psychology and neural science at New York University, argues that we could use notions from computer science to advance neuroscience. In a paper that was published in Science in 2014, he and his colleagues argued that a specific type of computer, a field programmable gate array, could function similarly as the brain.
In particular, this computer works as a set of reprogrammable building blocks that can take on different tasks. For example, as Marcus explains, one block could be in charge of vision, another would do arithmetic and another would process signals. They suggest that the brain may be similarly structured in elemental basic blocks. And that these brain primitives are what we should look for to understand the brain.
The opposite approach would be to try to map every brain process, from neural activity to human behavior, but Marcus criticizes this idea. As he puts it:
“It is unlikely that we will ever be able to directly connect the language of neurons and synapses to the diversity of human behavior, as many neuroscientists seem to hope.”
He argues that finding a robust middle ground is necessary to go from understanding the most basic level of brain processing to understanding the most complex and that those mental building blocks may hold the key for it.
We’ll understand the brain by looking at the brain
But some refuse the idea that the brain is a computer. Not just because they think it’s not true, but because it has no use in advancing our understanding of the brain.
There are some strong arguments against the computational notion of the brain, in particular against the most basic aspect of the metaphor: that the brain has a neural code. That is, it internally represents the external stimuli.
Instead, what we know is that there is a relationship between stimuli and neural activity. But we don’t know whether this activity represents the stimuli or not.
György Buzsáki, professor of neuroscience at New York University, says that, while a computer is passively taking information and representing it into a neural code, the brain is part of a human being that actively interacts with the world. The brain takes information to then search for possibilities to make sense of it. As Matthew Cobb explains in reference to Buzsáki’s argument:
“His conclusion — following scientists going back to the 19th century — is that the brain does not represent information: it constructs it.”
The detractors of the computational analogy say that computers are not a good source to look for knowledge of the brain, but no one is sure what the best approach to study the brain is. Some prefer to focus on developing better mathematical models. Others say that it’s better to study the brains of smaller organisms, such as the worm C. elegans. Others think that studying simple processes is the way to eventually understand the emergence of complex phenomena, such as consciousness.
But, what they all agree on is that it’s by studying the brain that we will understand the brain.
So what can we get from this debate?
The most fruitful approach is to disentangle the metaphor that the brain is a computer without taking any side a priori. It may have some value, so we should try to assess what that value is and assess its limits and scope.
Interestingly, there have been other metaphors for the brain before. One of the first ones is the hydraulic theory of René Descartes in the 17th century. He suggested that the brain produced the movements in the muscles by flowing animal spirits through tubes inside the body.
Why did he think this? Because he was interested in hydraulics. And there have been other brain metaphors in history, each corresponding to some new technology of the time.
These metaphors provided some utility. For example, they served as analogies in which science could base new experiments. But they all eventually lose that utility when science advances enough to surpass the limits of the metaphoric scope.
Once we develop stronger theories for the brain , or make discoveries, or invent better methodologies, the metaphor of the brain as a computer will be of little use. And then, a now inconceivable technology may appear to relieve computers as the best comparative for brains and to serve as the next paradigmatic metaphor.
In conclusion, we could say that the brain is, and isn’t a computer. Because, as Matthew Cobb says, a metaphor is always only partial in nature.