Frontiers are in short supply. No explorer will again catch that first glimpse of the Pacific Ocean with “wild surmise,” take the first steps on the moon, or arrive first at the Challenger deep – the remotest corners of the earth are now tourist attractions. Even in science, great mysteries have fallen – life itself has gone from being the subject of metaphysical speculation about vital substances to the biophysical understanding of cellular processes. Uncharted territories, both physical and metaphorical, are hard to find.

Yet there is one largely unmapped continent, perhaps the most intriguing of them all, because it is the instrument of discovery itself: the human brain. It is the presumptive seat of our thoughts, and feelings, and consciousness. Even the clinical criteria for death feature the brain prominently, so it arbitrates human life as well. One would think, that after a century of intensive research, its outlines would be well known to us: after all, colorful pictures of brain activity have been making regular appearances in the news media for some time. 

However, if one scratches the surface, our knowledge of how the human brain is put together remains limited: not in some esoteric, complicated manner, but in the straightforward sense that we have simply no means to visualize entire neurons in the brain (and the brain, being a collection of neurons, therefore remains a shut book in important ways). We can’t see them in their full glory, even with all our advanced technology.

The problem is that compared to other cells visualized under a microscope, neurons are at the same time very small, and very big. While their soma (cell bodies) are like other cells, neurons can send out branches (axons) that travel very long distances, sometimes several feet, which don’t fit into the sections of tissue that we do histology on. We can see the distribution of neuronal cell bodies in a slice of the brain. And we can divide the brain up into regions based on their properties; this is what Korbinian Brodmann did a hundred years ago. Unfortunately, the most interesting properties of neurons come from their extremely long filaments, which can span the entire brain or reach down into the spinal cord. We cannot effectively label these in their entirety. It is as if we were presented with a map of a vast land with states and cities marked, but not the roads.

There do exist labeling methods suitable for (postmortem) human brain tissue: the Golgi method, that sparsely labels neurons using a silver precipitate, or the passive diffusion of lipophilic dyes dissolved in neuronal membranes. However, these methods give us pictures of small portions of the full spatial extent of neurons in the human brain, or pictures of thin sections of brain, with pieces of the neurons in them. No one has yet seen, under the microscope or in digital reconstruction, a complete human brain neuron that sends projections to distant parts of the brain. To do that at the whole-brain scale, would be like seeing a new continent or planet.

In the live human brain, the methods at hand are even more indirect: rather than visualizing neurons directly, one observes the diffusive motion of water protons, using variants of magnetic resonance imaging. The basic idea behind these methods is to watch the diffusive spread of water protons starting from a point. If the diffusion was unrestricted, for example in a glass of water, the spread would be equal in all directions. If however diffusion was easier in some directions (eg parallel to a bundle of axons) and harder in other directions (perpendicular to the bundle) then one could figure out which way the fibers were running just by observing the diffusion of the water protons. This is the methodology used in a recent paper by Van Wedeen et. al., which has excited much attention.

The authors claim that there exists a surprising and pervasive degree of geometric regularity in the long-range neuronal trajectories in the brain. They report that throughout the brain’s white matter is to be found a grid-like pattern of fibers crossing at right angles. The brain’s superhighways, the authors tell us, form a three dimensional analog of the streets and avenues in New York City, running at right angles to each other. Note that this does not tell us about individual neurons: even if neuronal processes group together on grid-like highways during part of their journey across the brain, that does not tell us how different parts of the brain are connected (just as one needs to know more than which highways to travel on when visiting a friend living in a different city). Nevertheless, for something as complicated as the human brain, all regularities are welcome news.

It is worth examining critically the novelty and nature of these striking claims. The paper offers no numerical quantification of the prevalence of these putative grid-like patterns in the brain (eg, what fraction of the white matter shows grid patterns – or, what is the degree to which such a pattern occurs at some given point of the brain), but contents itself with attractive visualizations. For example, we think of the Corpus Callosum as running laterally, joining the two brain hemispheres; to what extent are fibers running front-to-back, or up-down embedded into the Corpus Callosum? It is also important to remember that the method is indirect, and the results have not been validated using the “ground truth” of postmortem neuroanatomy. That there are regularities in fiber pathways of the brain, and even the qualitative presence of criss-crossing patterns, was previously known from myelin stained brain sections: what this current paper quantitatively adds to that prior understanding, is not yet clear.

In fact, bundles of fibers, running at right angles to each other in three orthogonal directions, were already described more than a century ago, as can be seen in this illustration from “Anatomie des Centres Nerveux” (1895) by Joseph Jules Dejerine. This Weigert stained section of the human brain shows neatly organized fibers of the Cingulum (horizontal), the Corpus Callosum (vertical), and the Corona Radiata (Couronne Rayonnante in the French text) perpendicular to the page. We haven’t come that far from Dejerine as far as the wiring of the human brain is concerned: this is what the late Francis Crick and Ted Jones, termed “The Backwardness of Human Neuroanatomy.”

We need a real technical breakthrough that allows us to trace neurons across the postmortem human brain. Meanwhile, following Darwin, we can learn something about human brains by studying non-human animals. Even there, we suffer from a substantial knowledge gap. The complete genomes of multiple species have been mapped, but we don’t have similarly complete maps of brain circuits. Recently, I joined with colleagues in calling for an effort to close this gap with systematic circuit mapping projects in multiple species. I am engaged in such a project myself to systematically map the circuits of the mouse brain using histological methods, and related efforts are under way elsewhere. These are early days: to get at the full complexity even of mouse brain circuits will require significantly more resources than we are devoting to it now. To do something at the same scale for the human brain is a challenge that is as daunting today as it was to set sail for an unknown continent, hundreds of years ago. There are still some real frontiers left for us to breach.

Are you a scientist who specializes in neuroscience, cognitive science, or psychology? And have you read a recent peer-reviewed paper that you would like to write about? Please send suggestions to Mind Matters editor Gareth Cook, a Pulitzer prize-winning journalist at the Boston Globe. He can be reached at garethideas AT or Twitter @garethideas.