Today’s Chronicle of Higher Education has an interesting story about connectomics—the study of our brain’s wiring diagram—and some of the fringe science like mind uploading that has grown out of the belief that this is finally the big idea that will crack the code. The trouble is: it isn’t, and it won’t. So please don’t run out and deli-slice your brain in the vain hope of waking up in a brand-new robot body. Because you’ll be disappointed. Or, actually, you won’t be; you’ll be dead.
Don’t get me wrong. I’m a huge fan of connectomics, and believe that it is going to teach us an immense amount the brain. But I’m not a fan of brain GUTs—that’s Grand Unified Theories of brain function. There isn’t going to be a single organizing principle for brain function, and we should stop looking for one. The brain functions as the result of the dynamic interaction between many different systems, at different levels of organization, each operating according to its own rules. Consider just three extra-connectomic contributors to brain function, at three different levels of organization:
1. Volume transmission. Also known as non-synaptic diffusion neurotransmission, this refers to a class of neuro-modulatory mechanisms that rely on the diffusion of molecules to cells not syaptically connected to the releasing cell. Classical wiring transmission relies on synaptic contact and diffusion barriers that keep the neurotransmitters relatively contained; diffusion is in this sense the enemy of reliable synaptic transmission. But there is also a kind of chemical communication in the brain that takes advantage of diffusion, and in particular of structural features of the brain that allow for greater diffusion in some directions than in others (called anisotropy), by releasing molecules from non-synaptic sites. Considering that pyramidal cells, for instance, typically synapse with only about one percent of the cells within their dendritic arbor, the functional potential of volume transmission is enormous. Volume transmission is known to play a role in learning and to modulate synaptic transmission, and may also play an important role in generating new long-distance partnerships in the brain. To say that volume transmission is poorly understood would be a serious understatement, but the point here is that this aspect of brain function cannot be captured by connectomics even in principle.
2. Neuron-glia interactions. There is a growing body of evidence that glial cells—what R. Douglas Fields has called “the other brain”—are more than just the nutritional and structural background to neurons. Instead, they make important contributions to brain function. Glia are thought to regulate the formation of synapses, modulate learning mechanisms such as long-term potentiation, and regulate synaptic transmission because they both manage the clearance of neurotransmitters from the synaptic cleft, and also release their own neuro-modulatory substances. Fields writes: “Glia communicate, and they provide a separate, non-electric network for information flow through our brain that operates independently but cooperatively with neuronal networks. Glia do not use electricity to communicate, so glia have no need for the axon or dendrites or synapses of nerve cells. Glia communicate by broadcasting chemical messages. Moreover, glia can sense information flowing through neural circuits and alter the communications between neurons at synapses!” None of this crucial interaction is captured by connectomics.
3. Embodiment. The brain evolved first and foremost as a control system for a particular kind of organism—you—in a particular kind of environment. Human cognition is not just human brain function, but human brain-body-environment function. It is marked by the use of and interaction with the environment in myriad ways: using a pencil and paper to store intermediate results in long division or large-number multiplication; arranging a hand of cards or scrabble tiles to better see relevant patterns, matches, or potential words; rotating puzzle pieces to better discern their fit; making grocery lists, labels, signs, encyclopedias, and otherwise storing information in the world to be consulted later; and using management structures, and the constraints imposed by individual social roles, to accomplish complex tasks like ship navigation or building construction. The overall picture that this suggests is of an intelligence that lies less in the individual brain—and even less in that brain’s internal connections—and more in the dynamic interaction of agents with and within the wider world. Mindedness at this level of description is the activity of making the world a home, a home that reflects the nature of its occupant. Its primary sign is a kind of adaptive integration with the environment, including especially the social and cultural worlds that are so important to human cognition.Taking connectomics as a brain GUT is, in contrast, a return of an odd kind of Cartesianism.
So the connectome is important—crucial in fact. But so are volume transmission, neuron-glia interactions, and embodiment. And I haven’t even mentioned hormones, or local and large-scale oscillatory dynamics, or intercellular proteins, or many of the other important structures and processes that are known to make crucial contributions to brain function. So, definitely let’s study brain tissue using all the means at our disposal; but let’s leave brain GUTs alone.