Excerpt: 'An Internet in Your Head' by Daniel Graham

May 3, 2021

Andreas Vesalius, from De Humani Corporis Fabrica (On the Structure of the Human Body), 1543

What Good are Brain Metaphors?

We think of our brains as little computers, busily tallying up bits of information. We “retrieve memories,” we try to “reboot” or “upgrade” our minds. Neuroscientists were the first to use this metaphor and it has propelled countless discoveries. It still guides most of our theories and experiments. Yet as I argue in An Internet in Your Head, it is time for a new metaphor for the brain: the internet.

But why do we need any metaphor for the brain? Fields like chemistry don’t seem to require guiding metaphors. The idea of an acid or a base isn’t a metaphor: it’s a physical property of a solution that can be measured precisely in terms of pH; the terms acid and base are just shorthand. Why does biology in general and neuroscience in particular seem to need metaphors, often technological metaphors? Is this a good thing?

The short answer is that, in the physical sciences, theorists use models that are very precise metaphors. The theory behind physics, for example, is at its most basic a toy universe that we construct in our heads and express in mathematics. But the toy we imagine behaves just like the world in all observable ways, and it predicts real-world events very well. No one has ever observed an electron spinning. Yet physicists treat an electron as a kind of imaginary toy—a tiny bar magnet that spins (albeit in quantized fashion). This model successfully predicts everything about the electron that we can measure, like its interactions with other particles. Scientific theory is not identical to the world it describes. If it were, it would be useless. Its job is to expose mechanisms of action and their origins.

A metaphor in science is simply a vague theory, or the beginning of a theory. It is most useful if it is deep. The target of the metaphor should have many layers that are formally articulated and can be compared in detail to the system under study. The deeper the metaphor, the closer it is to theory. The goal is to capture key aspects of the system.

Metaphors are unfamiliar in the physical sciences because the reigning metaphors are precise enough to be couched in terms of mathematics. But metaphors—and especially technological metaphors—have been critical in the history of science, and they will continue to be so as we get closer to understanding the brain.

The Enlightenment of scientific thought that started in the seventeenth century depended heavily on the innovative technology of that era. But the engineering of the early Industrial Revolution also gave us metaphors. Most importantly, it gave us the notion of technology itself.

The word technology, though borrowed from ancient Greece, was first used in its current meaning in the seventeenth century. Linking the world’s workings to technology implied that the universe could be a mechanism like a gloriously complex mechanical clock, one perhaps fashioned by God. Mechanisms, unlike metaphysics, could be studied through experimentation because they worked according to fixed, observable rules. All we need to do is to imagine a possible mechanism in the first place, then look for evidence of it. Reductionist science from then to now has largely been built on this premise.

Technological metaphors have been particularly important in neuroscience. As with the physical sciences, it was the earliest and most basic technological metaphor that has been most important for understanding the brain, and for the same reason.

René Descartes was the first to explicitly build a metaphorical linkage between natural phenomena and the hi-tech engineering of his day. In the mid-seventeenth century, he introduced the idea that living systems are mechanical automata. Descartes argued that each organism ran according to fixed rules. This insight opened new theoretical approaches, especially those that discounted the role of divine or celestial forces. If an animal is a complicated mechanism that runs on its own, the divine is less salient.

Famously, Descartes could not fully dispense with divine forces in biology. His distinction between the mystical human mind and our physical body bears his name today: Cartesian dualism. Descartes could tolerate mechanical rabbits and dogs, but humans and their seemingly special form of consciousness still required the supernatural. For better or for worse, Cartesian dualism and its implications would shape brain science through to the present day. Some have argued that what we call “cognition” today is nothing other than the mystical mind of dualism. But at least in terms of the “body” half of the dualism—including the brain—Descartes’s basic metaphor holds up today.

Within the broad metaphor of animals as mechanical automata, Descartes imagined a more specific metaphor for how the brain works. It was centered on plumbing (hydraulics). Plumbing is so pedestrian today as to be invisible. It seems to have existed forever. But even in 1940, almost half of all U.S. households lacked full indoor plumbing. In Descartes’s time, grand waterworks like those at the Palace of Versailles were a major engineering advance, one available only to the richest. The acres of water gardens at Versailles were constructed mostly in the second half of the seventeenth century for Louis XIV of France. Water was delivered and distributed over 35 kilometers of piping and artfully expelled in arcs reaching several meters. All the water had to be moved uphill more than a vertical half-kilometer from the River Seine; 250 pumps drew water from the Seine, powered by the river’s current. The pumps supplied more water than was delivered at the time to all of Paris.

In Treatise of Man, Descartes writes that our volitional minds work “just as you may have seen in the grottos and fountains of our King [Louis XIV], in which the simple force imparted to the water in leaving the fountain is sufficient for the motions of different machines, even making them play musical instruments, or speak words according to the diverse disposition of the tubes conducting the water.” Descartes highlighted a part of the brain we now call the pineal body as the master valve of this plumbing system. The pineal body is unpaired, meaning it is a single, central mass, rather than a pair of mirror-image structures on either side of the brain. The pineal body is located adjacent to one of the brain’s large fluid-filled ventricles. Today we know that the job of the ventricles is to nourish and protect brain tissue with cerebrospinal fluid. With his keen knowledge of brain anatomy, it is easy to see how Descartes would make the connection to plumbing in the pineal body and its neighboring ventricle. But the proximity of the pineal body to the brain’s plumbing is just a coincidence; the ventricles don’t control thinking or behavior.

Descartes’s model was incorrect, but elements of the plumbing metaphor were retained and reappropriated at the founding of modern neuroscience. Charles Sherrington, winner of the 1932 Nobel Prize in physiology or medicine for his studies of reflex responses, described neurons as “valve-like” structures. Neurons indeed regulate the flow of molecules in fluids, including neurotransmitter molecules and ions. Another founder of modern neuroscience, Santiago Ramón y Cajal, invoked the idea of a valve in the same context. In classical as well as contemporary models of neuron firing, the end result of an electrical excitation in a neuron is a “puff” of neurotransmitters—a spritz pumped into the synapse, not unlike the expression of a Versailles fountain. Modern neuroscience also recognizes the role of ion channels as a literal valve whose hydraulics concern charged ionic particles.

A metaphor can open new modes of conceiving the mechanics of a system beyond those already understood. Technology is an especially good target for brain metaphor. In part, this is because technology is cumulative, meaning that innovation builds on itself, just as in living things. More importantly, the brain, like technology, has a purpose—or many purposes. Metaphor provides not only a mechanism, but also a goal.

Metaphors allow us to see the “hacks” or tricks behind a system. The brain has many tricks. Thanks in part to Descartes, we know that one of these tricks is plumbing (though Descartes was wrong about what exactly the plumbing system does). Another trick the brain uses is computing. I argue that internet-like communication is another. There are surely others. Each of these tricks resembles its engineered counterpart to a greater or lesser extent. We can learn much—and peer through a new theoretical lens—whenever we imagine an appropriate metaphorical linkage to one of these engineered systems. Plumbing systems and the brain differ in countless ways, but the primary trick behind plumbing is indeed one of the tricks behind brains passing information. Plumbing doesn’t explain everything. That’s why we need other metaphors.

The dominant metaphors for the brain are powerful systems of thought and language that allow us to analyze our measurements in a coherent way, even when we are dealing with highly formalized experimental procedures and fine-grained data. But this power comes with a cost. It imposes a cognitive frame on our understanding of the brain, both in science and in society. We can disclaim the computer metaphor again and again, as many neuroscientists do, but unless we augment it with something else, we are liable to return to its logic, even unconsciously.

At this point you might say we don’t need another bad analogy; we just need to look at the biology. An ion channel is not just a valve; it is also a dynamic biochemical system whose thermodynamics and genetic basis can be investigated. There is no question that basic biological processes can and should be investigated in the absence of metaphorical guidance. But an understanding of the complexity and interconnection of the larger system of a brain benefits from analogical thinking. Even a flawed analogy can spur new questions and lead to better scientific understanding, compared to an approach that supposes only an assemblage of specialized phenomena.

In some ways, it is odd that it has taken so long to recognize that flexible, efficient, reliable communication is precisely what both the internet and the brain do. The key innovations—such as a system for dividing up messages into chunks of fixed size—have been known for more than half a century. We should not expect that the brain works exactly like the internet. But tricks similar to those used by the internet seem necessary in the brain as well.

We still need metaphors for the brain because we are so far away from understanding the brain. We also still need metaphors because the brain does many different things. But even if we wanted to dispense with metaphor we couldn’t, because the brain is each one of us. We identify very strongly with it, and we need a way of understanding it. Metaphor can help.