Anatomy is the handmaid of physiology. From aardvarks to zyzzyva, if you seek to understand how something works, you must first tally its parts. Those who would take on physiology of the neural kind face a particularly uphill battle. First, brains are not built from the kind of detailed bauplan that applies to the body writ large. The shinbone may be connected to the knee-bone, but no such explicit instruction set exists for placing billions of neurons and trillions of synapses amongst them. There simply isn't enough information in the genome. Instead, we make do with statistics. Nature sketched some rules -- more what you call guidelines than actual rules -- threw us into the stew, and then washed her hands. The statistics clearly work (else you not reading this would be) but that does not mean sussing them is an easy task.
Second, even though we desire statistics writ large, it must be arduously written from the small. Neurons = tiny. You can't even see the pesky things without a microscope and even then not without preparation. Leeuwenhoek would have to wait a hundred years for the histological stains that make neurons appear in his invention, and a further hundred years for the electron microscope necessary to visualize synapses. Functional neuroanatomy is a Herculean census not of citizens but of invisible specks. No wonder protoscientists settled on the various lumps and gobbets as the fundamental units. Had you informed Descartes that it is not the pineal itself but a vast winking population of invisible animalcules within that comprise the soul you would have blown his mind.
Enter Valentino Braitenberg and Almut Schüz. Not household names (such accolades are heutzutage reserved for the sportsballer and "rappers"), but names not unknown to the neuroscience community. They spent five decades staring through microscopes at all things brain. Reporting what they saw, drawing what they saw, and -- most importantly -- quantifying what they saw. CORTEX: Statistics and Geometry of Neuronal Connectivity is a summary of those labors, and one of the few books on cortex that leaves the reader hopeful that, yes, it will be possible to one day understand its function and so understand ourselves. (At the other end of the optimism spectrum sits Peters and Jones Cerebral Cortex, currently at fifteen soul-crushing volumes and counting. The map is not the territory fellahs, Hayakawa would scold.)
In brief, C:S&G is an altar of neurostatistics. There are chapters on numbers of neurons, number of synapses per neuron, dendrite length, axon density, laminar distributions, and much more besides, along with occasional diversions detailing the various corrections and mathegonics such work entails. The latter chapters flirt with Hebbian cell assemblies and synfire chains and the final wanders into speculation about language processing and is probably more convincing if you studied the original articles and not an eight-page summary. But so what. If anyone has earned the right to a little speculation it's these guys.
Comparisons to Ed White's Cortical Circuits are unavoidable but Braitenberg is I think the better read. I find White's presentation overly obsessed with receptive fields -- which I personally don't think is a useful way of thinking about neurons -- but that's more the fault of Hubel and Wiesel. White also recognizes a number of neuron types that B&S do not. Go figure. However, both books deserve a place on your shelf. These were masters in their field writing at the top of their game.
Currently rather pricey in paper form but the Kindle is reasonable if you're into that sort of thing. Amazon click here making doing.
Second, even though we desire statistics writ large, it must be arduously written from the small. Neurons = tiny. You can't even see the pesky things without a microscope and even then not without preparation. Leeuwenhoek would have to wait a hundred years for the histological stains that make neurons appear in his invention, and a further hundred years for the electron microscope necessary to visualize synapses. Functional neuroanatomy is a Herculean census not of citizens but of invisible specks. No wonder protoscientists settled on the various lumps and gobbets as the fundamental units. Had you informed Descartes that it is not the pineal itself but a vast winking population of invisible animalcules within that comprise the soul you would have blown his mind.
Enter Valentino Braitenberg and Almut Schüz. Not household names (such accolades are heutzutage reserved for the sportsballer and "rappers"), but names not unknown to the neuroscience community. They spent five decades staring through microscopes at all things brain. Reporting what they saw, drawing what they saw, and -- most importantly -- quantifying what they saw. CORTEX: Statistics and Geometry of Neuronal Connectivity is a summary of those labors, and one of the few books on cortex that leaves the reader hopeful that, yes, it will be possible to one day understand its function and so understand ourselves. (At the other end of the optimism spectrum sits Peters and Jones Cerebral Cortex, currently at fifteen soul-crushing volumes and counting. The map is not the territory fellahs, Hayakawa would scold.)
In brief, C:S&G is an altar of neurostatistics. There are chapters on numbers of neurons, number of synapses per neuron, dendrite length, axon density, laminar distributions, and much more besides, along with occasional diversions detailing the various corrections and mathegonics such work entails. The latter chapters flirt with Hebbian cell assemblies and synfire chains and the final wanders into speculation about language processing and is probably more convincing if you studied the original articles and not an eight-page summary. But so what. If anyone has earned the right to a little speculation it's these guys.
Comparisons to Ed White's Cortical Circuits are unavoidable but Braitenberg is I think the better read. I find White's presentation overly obsessed with receptive fields -- which I personally don't think is a useful way of thinking about neurons -- but that's more the fault of Hubel and Wiesel. White also recognizes a number of neuron types that B&S do not. Go figure. However, both books deserve a place on your shelf. These were masters in their field writing at the top of their game.
Currently rather pricey in paper form but the Kindle is reasonable if you're into that sort of thing. Amazon click here making doing.
No comments:
Post a Comment