Connectome: How the Brain's Wiring Makes Us Who We Are

The wiring between individual neurons in your brain, your "connectome", determines the flow of neuronal activity, which ultimately produces all your perceptions, emotions, conscious thoughts, memories and behaviour. "You are your connectome": this is the central idea of Sebastian Seung's book. You might already agree that "you are your brain", but Seung's claim is stronger. He dares to propose that the dynamics within each individual neuron and other brain processes, such as the diffusion of neuromodulators outside specific synapses, may not be essential to the dynamics that define our personality and consciousness. It's the connections that matter. Understanding how neurodevelopmental disorders such as autism and schizophrenia and neurodegenerative disorders including Alzheimer's and Parkinson's affect the connectome might help us treat them with drugs and behavioural training.

Connectomes can be defined at multiple levels of organisation to describe the brain. Seung credits the term "connectome" to Olaf Sporns, whose work focuses on a coarser level of description, that of brain regions. While magnetic resonance imaging can reveal brain anatomy, connectivity and activity at this coarser scale, the neuron-level connectome that Seung prefers requires post-mortem slicing-up of our brains for automated microscopy and computer reconstruction of the detailed anatomical structure. This level of description, he argues, is special in that it might explain the mind.

As a neuroscientist, I'm tempted to quibble. Is the neuron-level connectome, Seung's preferred level of description, really so special an abstraction as to deserve the monumental scientific effort (and grand-scale funding) required to capture it? It might be missing essential information if the workings of our minds depend on the unique dynamics within each neuron or on diffusion processes outside the connectome. At the same time, the gigantic amounts of data that would describe the neuron-level connectome for a whole human brain (one cubic millimetre of mouse brain is where we are today) might have a lot of unnecessary anatomical detail. When the goal is to capture brain function, it might be better to abstract from the precise shape and connections of every tree in the neuronal forest. Moreover, the much coarser level of description accessible to MRI appears more relevant to medical applications. After all, MRI can be used in living patients to image activity as well as anatomical structure, while Seung's approach means slicing up their brains (a process no one has yet survived) and provides only structural information (albeit in stunning spatial detail). The coarser level of description provided by MRI is also easier to learn from, because statistical inference becomes notoriously challenging when the data get too detailed. Seung's enthusiasm for the neuronal-level connectome is exciting and infectious. However, neuroscience needs to continue to rely on multiple levels of description.

Seung's book is deliberately subjective in style and beautifully written. At the same time, it is careful enough in its argument to make for a compelling read. It will be exciting as a lay reader's first foray into neuroscience. I also recommend it to neuroscientists who will value Seung's coherent perspective on where we are today and his inspiring vision for the future. The first half of the book is a loosely connected guided tour of some of the highlights of the history of neuroscience. In the second half, Seung makes his heroic leap of faith, proposing that the dead structure of the brain, captured at the level of the neuronal connectome, will allow us, finally, to understand our unique minds. The book is brilliant in the way it brings the vastly complex but utterly dead structure of the sliced-up connectome to life in our minds. And, quibbles notwithstanding, it makes a good scientific case for slicing and computer reconstruction as one route towards understanding the brain.

The final two chapters examine "cryonics" and "uploading". Cryonics is the freezing of the brain (and optionally of the body) for later resurrection. Uploading is the more radical vision that our brains, once sliced and scanned, might be reanimated in a computer simulation, thus resurrecting our consciousness "in the matrix", where our virtual bodies and minds may live on indefinitely. This has the added benefit of improvements, potentially fulfilling Radiohead lyricist/vocalist Thom Yorke's poignant wish: "I want a perfect body, I want a perfect soul." Both cryonics and uploading depend on a future world that has developed the necessary technologies. If Seung is correct that the connectome is the right abstraction, then slicing up the brain for scanning and reconstruction in a computer may be the way to save it from the inevitability of death. It might well be the only way.