Minds and Machines

, Volume 23, Issue 2, pp 259–262

Olaf Sporns: Networks of the Brain

MIT Press, Cambridge, MA, 2011, xi+243, $40.00, ISBN 978-0-262-01469-4


    • Tilburg Center for Logic and Philosophy of ScienceUniversity of Tilburg (NL)
Book Review

DOI: 10.1007/s11023-012-9294-y

Cite this article as:
Colombo, M. Minds & Machines (2013) 23: 259. doi:10.1007/s11023-012-9294-y

Consider the World Wide Web, the banking system in Europe, a pack of wolves in Alaska, the Italian football federation, a criminal organization in Brazil, and the human brain. What do they have in common? Here is an answer: They all comprise intricately interconnected elements whose organized activity gives rise to peculiar phenomena and behaviour. In short, they are all complex networks.

Over the last decade or so, research on complex networks has expanded across diverse fields in the social and natural sciences alike, exploring the behaviour of complex systems ranging from cells to economic institutions. Networks of the Brain is a passionate exploration of the myriad ways in which a network approach can be brought to bear on neuroscience, so as to re-shape the methodologies that are used to study the brain as well as the questions we ask to understand how the brain works.

With this book, Olaf Sporns wants “to introduce networks to neuroscientists and make neuroscience appealing to all those working on theoretical network models” (p. ix). Networks of the Brain is then a book about the encounter between two worlds: the world of network science and the world of the brain sciences. This book, however, is neither a neuroscience textbook nor a textbook in network modelling.

Sporns’s aim is to highlight the network “revolution” that is taking place in cognitive neuroscience. This revolution consists in the discovery that tools and concepts from network science can empower the methodological arsenal with which fundamental questions about cognitive architecture, functional specialization and integration, mental abnormality, individual differences in psychological profile, and about the roles of body and external environment in cognitive processing can be fruitfully tackled. This network revolution recommends a shift in perspective when we formulate the questions directed at understanding how the activity of the 1011 and more cells that constitute the human brain, with their 1015 and more connections, gives rise to our mental life. It commends us to recognize that fundamental principles of brain architecture and function can be illuminated from a complex network perspective. Such a perspective promises to bring under a unifying umbrella otherwise scattered phenomena and data in the cognitive neurosciences, because, after all, “it’s networks all the way down”—as Sporns explains (p. 325). We are in fact embedded in social networks, which affect who we are and how we behave, our genes interact in gene networks, and the proteins, which are the building blocks of our organisms, interact in signalling networks.

Given these goals, it should be clear that Sporns’s book will be of broad appeal among neuroscientists, cognitive scientists, philosophers, and anyone interested in what network science can contribute to our understanding of the brain, cognition and behaviour.

Networks of the Brain is systematic, rich, balanced and accessible. The book is in fourteen chapters. Each chapter has as a starting point a quotation by one of the “hub-figures” in the development of modern networks neuroscience, from Leonhard Euler (the grandfather of network science) to W. Ross Ashby and Alan Turing. This is an engaging feature of Sporns’s book, which shows sensitivity to the historical roots of the novel field he is mapping out.

The first two chapters offer a survey of the basics of network science (Chapter 2) and a description of some of the main techniques and approaches, along with the methodological and conceptual issues they give rise to, used to identify and analyze brain networks from raw neuroscience datasets (Chapter 3). In spite of introducing technical terminology and a variety of methods from graph theory, statistics and neuroscience, these two chapters contain—as the rest of the book—not a single equation. For the interested reader, nonetheless, Sporns provides plenty of references to more advanced resources in network science modelling and its many applications.

The jargon and the level of detail in introducing some of the concepts and techniques might be off-putting for some of the readers; yet, Sporns illustrates such concepts and techniques with a number of worked-out case-studies in an effort to make them digestible for everybody. Furthermore, a “Network Glossary” at the end of the volume explains clearly and concisely the key terminology from network science used throughout the book.

After having set the stage, Sporns proceeds to put network ideas and methodology at work. He begins to articulate one of the key insights of his book: we cannot fully understand how and why the brain carries out certain cognitive functions unless we understand the constraints and functional opportunities afforded by the idiosyncratic patterns of structural connectivity that link neural elements at multiple spatial scales.

Chapters 4, 5 and 6 pursue this insight by focusing on how functional specialization is related to anatomical networks from the microscale of individual neurons and synapses to the mesoscale of populations of neurons and brain regions, to the macroscale of embodied cognitive systems interacting with their environment. Chapter 6 deserves particular attention. It explains how and why the architecture of the brain follows a “small-world” topology, in which connected nodes have highly overlapping sets of partners, and yet pairs of nodes are, on average, connected via short paths (that is, the number of edges linking any pair of nodes is small). This means that, while the brain exhibits dense patterns of connectivity between neighbouring neurons, it presents also relatively few long-range connections. As a consequence, its connectivity pattern is locally dense and globally sparse: the brain network thus follows a small-world topology, which has crucial effects on its functionality. This type of architecture is important because it saves wiring space and promotes efficient information processing as well as diverse and complex network dynamics.

The treatment here is balanced, and supported by many case-studies and figures. Sporns rejects a reductive approach to explanation: he maintains that the existence and character of phenomena at a given level of neural organization cannot be made intelligible only on the basis of phenomena and regularities at some other, allegedly privileged level. He also makes clear that, in spite of much excitement for the connectome project—that is, for the project of drawing the complete wiring map of the brain (by the way, I suggest the interested readers to watch out for Sporns’s forthcoming book specifically on the connectome)—we should not forget that, by itself, a complete wiring map of the brain does not suffice to uncover the principles underlying the complexity of human cognition. We must investigate maps of the brain guided by knowledge of the biophysical details of neurons and synapses as well as by more abstract principles, drawn from such fields as evolutionary biology, dynamical system theory, developmental biology and information theory. In this respect, Sporns is a lucid advocate of explanatory pluralism in cognitive neuroscience.

Such pluralism is apparent in the three chapters that follow. Chapter 7 explores the question of what drives the evolution of the architecture of brain networks. Rather than asking whether the structural design of the brain is “optimal,” Sporns urges us to pay closer attention to the many factors that shape the evolution of the brain architecture. Chapter 8—one of the most intriguing—is primarily devoted to network dynamics generated endogenously, from spontaneous activity, and their functional significance. Spontaneous neural activity should not be understood as mere “noise;” rather, it might support and maintain an internal functional repertoire, which might serve as a background of expectations that guide sensory processing and adaptive behaviour. Chapter 9 is also central. It explains the claim that cognition “is a collective property of very large numbers of neural elements that are interconnected in complex patterns” (p. 181). It breezes through such topics as cognitive modularity, hierarchical processing and individual differences in a way that promises to advance and reorient traditional debates within theoretical cognitive science and the philosophy of psychology. One such debate concerns modularity of mind, another concerns the “binding problem.”

In relation to the binding problem, that is, the problem of rapidly and coherently unifying specialized and distributed information, this chapter articulates one of the motifs of the book: the relationship between functional segregation and integration. Segregation indicates that neural activations represent specialized information; integration of widely distributed and specialized resources is required for coherent cognition and adaptive behaviour. Sporns does a great job in showing how architectural principles of the brain network can solve the binding problem, supporting the opposite computational demands of functional segregation and integration.

A network perspective might also be fruitfully brought to bear on mental illness. Chapter 10 draws such a link, suggesting that psychiatric conditions such as Alzheimer’s disease, schizophrenia and autism can be understood as specific brain network disruptions. Chapter 11 integrates and expands on themes already tackled in Chapters 7 and 9. It focuses on the development of biological neural networks, with an emphasis on how networks’ dynamical differences across life-spans relate to cognitive and behavioural variation.

In the final three chapters, Sporns completes his journey by focusing on different aspects of network complexity. Chapter 12 asks what are the dynamic processes sustained by multiscale neural network architectures and how they relate to efficient information processing. Chapter 13 clarifies what it means for the brain to be a complex system and why that matters. This chapter critically surveys different measures of complexity from an information-theoretic perspective, and relates them to the structural and functional organization of the nervous system. Having explained what complexity could be, the chapter briefly discusses why complexity can matter for understanding the nature of consciousness and what may drive the evolution of brain networks. Both sections are thought-provoking and deserve attention from scientists as well as from philosophers of mind and biology. The last chapter, Chapter 14, is one of the most engaging. It broadens the topic of the book by examining the role of the body and of the environment in sustaining the kinds of information processes that give rise to our cognitive life and adaptive behaviour. It explains with excitement what it means that “it’s networks all the way down” (p. 325).

A network perspective has already opened new research avenues in neuroscience. Sporns passionately argues that a network perspective also promises to illuminate many of the open questions about our condition as cognitive agents. A view from network science will help us recognize how intricate flows of information through dynamic, complex interconnected elements spanning multiple temporal and spatial scales—from genes to our great many cultural and social institutions—combine to make us human.

Networks of the Brain is a unique resource. It defines the nature and scope of one of the newest and most exciting research programs in cognitive neuroscience.

Copyright information

© Springer Science+Business Media Dordrecht 2012