Mycorrhizal networks

Are there any science books that you recommend? Yes, several. Shadows of Reality, by Tony Robbin (Yale University Press) is a beautiful book that addresses the problem of illustrating high-dimensional entities. The issue is central to the sorts of things we deal with in science today, especially so in my field. How do we think and make sense of abstract spaces? This book is a great introduction to the topic. It is beautifully illustrated — much like Edward Tufte’s books — but also filled with interesting maths, physics, arts and history. Sparse Distributed Memories by Pentti Kanerva (MIT Press) is a more technical monograph that talks about the structure and advantages of binary spaces. Kanerva is a computer scientist and his theoretical exploration arose from an interest in cerebellum architecture, very much like David Marr’s early work. A very interesting book to start thinking about memory and representations. Principles of Brain Evolution by Georg Streidter (Sinauer) is another. One cannot think about biology, or the brain, without considering evolution. I think that experimental systems neuroscientists are some of the biologists least ready to acknowledge the value of comparative approaches and the advantages of simpler systems. Our kind of neuroscience regrettably is a very parochial corner of biology. Reading comparative neuroanatomy reveals how much there remains to discover about brain evolution. The number of unresolved issues means that people fight over ideas. Reading Georg’s book is a treat, and a rare one with anatomy books. Walter Heiligenberg’s Neural Nets in Electric Fish (MIT Press) is a fantastic little book, published before Walter’s tragic death and sadly out of print when I last checked. It describes nearly all that was known about the jamming avoidance response of the electric fish Eigenmannia in 1990, and how one can go about understanding a neural system, from ecology to behavior, structure and computation. People call this neuroethology. I call it neuroscience, in its ideal form.