Customer Reviews

Linked: The New Science of Networks

5 star:		(61)
4 star:		(25)
3 star:		(14)
2 star:		(10)
1 star:		(6)

 

 

Updated 28 Dec 07 to add links.

I have mixed feelings about this book. On the one hand, it is coherent, thoughtful, and tells a story about the emerging science of networks that anyone, who can read, can understand. This is a non-trivial accomplishment, so 4 stars.

However, the book is also--being brilliantly designed to be understood by the lowest common denominator, an undergraduate--somewhat shallow and empty.... especially when compared with Stephen Wolfram's "A New Kind of Science", 1197 pages not counting the index, which is at the other extreme.

Although there are good notes, there is no bibliography, and the author fails to use network methodology to illustrate and document the emerging literature on networks--called citation analysis, this would have been a superb appendix to the book that would have taken it up a notch in utility.

Among the key points that the author discusses and which certainly make the book worth buying and reading, my above reservations not-with-standing:

1) Reductionism has driven 20th century science (and one might add, all other knowledge), with the result being that we have experts who know more and more about less and less--and )as CIA and FBI recently found)while leaving us devoid of generalists and multi-disciplinary artists and scientists who can "connect the dots" across these fragmented foci.

2) Contrary to the prevailing wisdom about networks being equally distributed and thus largely invulnerable to catastrophic meltdown, the author does a fine job of documenting the importance of selected "hubs", so important that their removal ultimately breaks the network down into isolated pieces. The functionality of the network, its strength, is also its weakness--vulnerability to deliberate attack against the hubs (the author does not mention the Internet domain directories except in passing while discussing a table error, but MAYEAST and MAYWEST would be two obvious directory hubs that could be better protected through replication).

3) The author inadvertently makes a vital contribution to our understanding of how to defend America against terrorism--discussing why no single authority can close down the Internet by fiat, he notes "The underlying network has become so distributed, decentralized and locally guarded that even such an ordinary task as getting a central map of it has become virtually impossible." LOCALLY GUARDED--this is the key phrase. Federalizing counter-terrorism, and using federal agents and computers at the state and local levels, will not be effective against terrorists in civilian guise within the homeland--only a complete extension of counterintelligence and counterterrorism methods to the state & local level--teaching them to fish for terrorists, rather than trying to catch the terrorists with federal trawlers, is the way to go.

4) The author flirts with what is known as nomadic computing, making the point that nodes built around individual people are becoming as important--some would say more important--in a networked economy than nodes built around static organizations. There is a useful general discussion of how "fitness" in a networked economy is a combination of speed and scalability as well as diversity of linkages. As a general rule, as the FBI found (and also CIA, INS, and the State Department), systems with a single hub resistant to initiative from the field offices will tend to be slow and ineffective.

Missing from this populist overview is a discussion of the vital importance of geospatial information. While the author helpfully notes the Earth is increasingly covered by an electronic "skin" with millions of measuring devices, with experts predicting that by 2010 there will "around 10,000 telemetric devices for each human on the planet" (one suspects this refers only to privileged humans, not the billions of dispossessed that lack telephones, never mind computers), he does not take the next essential step, which is to note that in the absence of an XML-GEO standard and a global push to associate geospatial as well as temporal tags with all data, much of what we collect will, like the trillions of bits we have collected with secret satellites, never get processed in a meaningful manner.

This is a helpful book that will be of value to the general reader at the elementary (adult) level.

See also:
The Tao of Democracy: Using Co-Intelligence to Create a World That Works for All
Group Genius: The Creative Power of Collaboration
Collective Intelligence: Mankind's Emerging World in Cyberspace
Global Brain: The Evolution of Mass Mind from the Big Bang to the 21st Century
One from Many: VISA and the Rise of Chaordic Organization
The Wealth of Networks: How Social Production Transforms Markets and Freedom
World Brain (Essay Index Reprint Series)
The Wisdom of Crowds
An Army of Davids: How Markets and Technology Empower Ordinary People to Beat Big Media, Big Government, and Other Goliaths

What do sexually transmitted diseases, the World Wide Web, the electric power grid, Al Queda terrorists, and a cocktail party have in common? They are all networks. They conform to surprising mathematical laws which are only now becoming clear. Albert-Laszlo Barabasi has helped discover some of those laws over just the past five years, and though they are some pretty abstruse mathematics, he has written a clear and interesting guide to them, _Linked: The New Science of Networks_ (Perseus Publishing). Not only has he attempted in this book to bring the math to non-mathematicians, he has shown why the work is important in down-to-earth applications.

It is important for those multitudes who have no taste for math to know that this is not a book full of equations; Barabasi knows that for most of his readers, doing the math is not as important as getting a feel for what the math does. He explains the basic history of network theory, and then shows how his own work has turned it into a closer model of reality, a model that most of us will recognize. Networks are all around us, and they are simply not random. Some of our friends, for instance, are loners, while others seem to know everyone in town. Some websites, like Google and Amazon, we just cannot avoid clicking on or being referred to, but many others are obscure and you could only find them if someone sent you their addresses. Barabasi calls these "nodes" with such an extraordinary number of links "hubs," and he and his students have found laws of networks with hubs, showing such things as how they can continue to function if random nodes are eliminated but they fragment if the hubs are hit. Barabasi is currently doing research to show what intracellular proteins interact with other proteins, and true to form, some of them are hubs of reactions with lots of others. Finding the hubs of cancerous cells, for instance, and developing ways of taking them out, show enormous promise in the fight against cancer. And finding the hub terrorists in Al Queda in order to take them out would be the best way to eliminate the network.

Barabasi obviously enjoys drawing examples from all over, and because of his ability to link them, his book is a pleasure to read. He also shows how this type of mathematics is being done, by conference in obscure European locales and by e-mail. He shows how "eureka" insights by his students have propelled the new science, and he is full of good stories from a teacher. In fact, he is a good teacher, and those who follow along here will have reason to be glad to join, if only in the role of isolated nodes, into this network of mathematical thought.

With so much buzz about Wolfram's book, great to see a book that DOES talk about NEW science. Barabasi, the top guy in the new science of networks, talks about what he knows best: complexity and networks, and how they affect our life. While an easy read, it is full of so many thought provoking ideas, that I'd read for a while and then have to put it down to reflect over the details of what I'd just read. Gladwell's tipping point was an entertaining read, but light on true understanding. Linked makes up the difference: it breaks new ground, offering the reader insight and research into the structure of networks in just about all fields and aspects of life. While Gladwell chats about connectors, people who are incredibly sociable and well-connected, Barabasi is the one who really gets to the heart of the matter. He discovered these connectors (he calls them hubs) while looking at the www (Yahoo and Google are some of those), and he shows that they are present in the cell, in the business world (Vernon Jordan), in sex (Wild Chamberlain), in Hollyood (Kavin Bacon) and many other networks. These hubs are not accidents, but they appear in all networks as a simple rich gets richer process is responsible for them.

If you REALLY want to grasp how ideas spread, how to stop AIDS, how to break down the Internet, how to use your neighbor's computer, how to make your website matter or how to became a board member in a big company, Linked is a good place to start. Barabasi breaks down a complex world into very simple, clear concepts. While I have read several books about 'new' science, this one is really about something new, exciting, and hard to forget. Highly recommend it.

I have focused this review on the audience of the book, since other reviews have quite adequately summarized the material.

There have been a lot of books recently that have been published on the new science of networks. Network theory and how it applies to many different fields from technology, marketing, biology, social science, terrorism, disease control etc. (Six Degrees by Duncan Watts, Nexus - Mark Buchanan, Smart Mobs - Howard Rheingold, Tipping Point - Malcolm Gladwell etc..).

Barabasi's is a welcome addition to the field and has a nice niche, which isn't filled by the other books. As some other reviewers have pointed this book is a popular science book, which means it covers scientific and mathematical theories at a very high level and makes these theories accessible to a wide audience. The niche lies somewhere between Gladwell's Tipping Point and Watt's Six Degrees. It is very well written and draws you in with stories that explore the theories. Some of the other reviewers have complained that Barabasi has done a disservice to the theories that he explains by making them too simplistic. I disagree, I actually found this book to be very rewarding, and a quick read, which is a sign of a well-written book. I have never been a fan of scientific and academic books that pride themselves on being totally incomprehensible. Richard Feynman, the Nobel Prize winning physicist, once said that if someone truly understands a subject they should be able to explain it to a general audience without resorting to technical jargon (Feynman's Lectures on Physics Vol 1,2,3 are a perfect example). To be able to explain a complex subject you need to resort analogies, examples and stories. Stories give a framework for the general reader to absorb the complex material. Barabasi has managed to explain the science of networks using all three. I am not sure how this can be seen as a bad thing. This exposes a wider audience to a very interesting subject; this has to be good thing.

Summary:
Anybody who loved Gladwell's Tipping Point and was looking for a book that explains some of the theories behind the phenomena will love this book. It's a little bit more technical than Gladwell's book, but it is well written and it will appeal to a wide audience. As popular science books go, this is definitely on par with Ed Regis's Nano and Steven Levy's Artificial Life, but not quite at the level of Gleick's Chaos. If you are looking for a technical book, you should look at Duncan Watt's Six Degrees, or Small Worlds.

Linked focuses on network theory and some of its applications, where networks are defined as dynamic linear graphs. It is written for the non-mathematicians, and in fact does a very good job of giving the reader insight into how the mathematical modeler thinks and works, and what mathematical modeling is (the phrase "dynamic linear graphs" does not actually appear in the text). "Linked" is kind of an intellectual memoir, and especially in the first few chapters, is charming as well as informative. The problem is that Barabasi has an inflated view of the importance and primacy of his work and interests vis-a-vis the general subject of the theory of complex systems. Also, while Barbasi strikes me as intellectually honest, his lack of knowledge of such subjects as cellular biology leads him to erroneous claims for what insights may be attributed to recent work in network modeling. He is on stronger grounds when he discusses narrow subjects such as the links between corporate directors, and Barabasi does seem to know quite a bit about sociological modeling and the Internet. In terms of intellectual stimulation and excitement, Linked does not begin to match up with Emergence: The Connected Lives of Ants, Brains, Cities, and Software by Steven Johnson, and I guess I was expecting something more comparable to that book.

This book is an absolute diamond.

I am not an explorer at the very frontiers of network dynamics, but I am an intelligent, sentient being and the ideas being developed in network research are of great importance to my life and the world I live in. I find the academic journals, where the research results are first published, (deliberately?) impenetrable, so it is a delight and a joy to find a guided tour in plain English, with an authoritative guide, through the frontiers of some very current and paradigm-changing ideas.

The book's narrative is aimed at the general public, to be sure, but I hadn't heard that being a member of the general public was a crime or a slur. Perhaps I missed a meeting. I found the writing style to be clear, concise, engaging and entertaining. In short, it was one of the best books of any genre that I have ever read and I have read hundreds.

Another reviewer of this book (see below) has said that Barabasi overemphasises the importance of preferential attachment in forming scale free network topologies. OK. Bring it on. Where is your counter explanation? What is the more important factor? Where is your clearly-written book explaining your counter argument for the likes of me? I would really like to know what else could possibly account for the emergence of this topology. It's important to me. To that reviewer I say, "publish or be dammed". I cannot abide elitism in scientific research, whereby those in the know jealously guard their secrets from the rest of us, so as to reinforce their self-belief in their uber-mortality. Join the real world. Tells us what you know without being patronising.

If, as a reader, you are in any way interested in the spread (or diffusion) of ideas, innovations, fads, viruses, memes, rumours and a hundred other phenomena or want to understand why some things are runaway hits and others not, this book will definitely stimulate your thinking. The only minor frustration I had with the book was that in identifying Microsoft's success in operating systems as analogous to a Bose Einstein condensate (a superfluid?), the book fails to explain how the condensate can evaporate...in other words, what nodal or network conditions would have to apply to overturn Microsoft's dominance?

As a published technical author myself, I know that writing this well is sheer hard work. I would be delighted if any of my own works were as brilliantly executed as Barabasi's "Linked".

Buy it. You won't be sorry.

Every good song has a hook--a line or a melody that makes you want you to hear more. For me the hook that kept me reading "Linked" appears on page 6. "Here is a secret that never makes the headlines" writes the author Barabasi, a professor of physics at Notre Dame, "We have taken apart the universe and have no idea how to put it back together. After spending trillions of research dollars to dissassemble nature...we are just now acknowledging that we have no clue how to continue -- except to take it apart further."

Could the prevalent reductionism be, at this point, obscuring rather that sharpening our vision of the the world? Should we give up on the idea of having an equation that could be printed on a tee shirt that contains the Theory of Everything?

Barabasi's answer is not to stop analyzing, but rather to analyze certain relatively large self organized systems topologically over time, as if you were reading successive maps of a growing cosmopolitan region. As roads, building complexes, public areas and the like are added, the rules generating the growth and distribution of facilities make themselves plain. The network thus formed works in a manner that obeys certain laws (called power laws)and divides into highly connected places called hubs which link to less numerously connected nodes. Using this analytical approach, he poses and answers dozens of questions that should stimulate an intelligent reader into seeing the world a little differently.

In fact, much of this book's interest derives from the activity of charting itself. Barabasi, like some latter day Henry the Navigator, explains in ordinary language why Kevin Bacon is really not very central to the Hollywood map and how the late Rod Steiger was. How Microsoft follows the same laws as a strange sort of matter. He explains how Vernon Jordan got so many corporate board seats, what may be the best way to eradicate AIDs, and what may become the newest phamaceutical approaches to disease. Terrorist networks, cancer and bad economics may all one day fall victim to strategies derived from network analysis.

I have read too many popular science books to know that translating equations into popular literature is no easy feat. So many fall of their very lack of weight. My one complaint with this highly readable book is that in an effort to make the book gallop the author has taken some shortcuts which break the flow of the argument. Nonetheless, if you want to read a very good book about a scientific issue not many of us are familiar with, this is a very good one.

Having read both "Linked" and "A New Kind of Science" I feel compelled to add my two cents to some other reviewer who unfavorably compares Barabasi to Wolfram.

While it is true that Linked is a bit light on the underlying math - not trivial by all means - and that there are chapters the book would be better without (last three notably, as well as the already-mentioned analysis of M$ dominance) this remains an interesting introduction to networks theory. We do not need rocket science to tell us that a scale-free network has its' vulnerability in its hubs, but I find it interesting and not entirely common sense that it is INHERENTLY more robust than a random network.

I find some of the critique here a bit petty (perhaps penned by fellow scientists ?). Barabasi comes out IMHO as a witted scientist with a knack for explaining stuff to the masses, an art in which Richard Feynman (alredy mentioned here and perhaps my all-time favorite hero) excelled. Perhaps a 100-page compendium would make a better reading, but there seems to be an unwritten publishing rule whereby no essay shorter than 250 pages sells.

On the other hand, I have rarely witnessed such an inflated ego as the one self-portrayed by Stephen Wolfram who bombastically claims to have invented a whole New Kind of Science ! His 1,200-page tome uses all variations of the "I" pronoun *ad nauseam* and there are whole sections who could be happily burned to no consequence to the reader (e.g. the proof-free wanderings on biochemistry et al.), not to mention the gazillion diagrams which cease to astonish well before you peruse the fiftieth.

Barabasi's Linked is a pretty good intro to the science of networks. It covers much the same ground as Buchanan's Nexus including discussions about random v. nonrandom networks, six degrees of separation related stuff, the AIDS epidemic, etc. Linked also covers a several more topics in greater detail than does Nexus including viruses and fads, and Barabasi presents very good discussions of search engines and the good-old-boy network of board members. Another appealing aspect of the book is that the author and his co-workers were involved in a number of the developments in the science of networks, so we can be sure that the author's explanations are grounded in his own experience.

If you want to read one book about networks and you're deciding between Nexus and Linked, I would recommend Nexus, however. Nexus' discussions are deeper, and its presentation and writing are better.

In all the chapters (author calls "links"- a novel idea for a book on networks!) Barabasi has nicely stitched the captivating overviews on different networking details and offer a general progressing view about network analysis. He used the "World Wide Web network" as a basis for network analysis and extended the findings to other possible real-world networks (cellular-proteins, diseases, network of actors at Hollywood, terrorist networks, web of scientists linked with co-authorship, collaborative web behind economy and so on).

As pointed in introduction, "reductionism" was the buzzword of twentieth century"s research based on the assumption "that once we understand parts, it will be easy to grasp whole" but the reassembly turned out to be harder leading us into "complexity". Barabasi projects the concept to the nodes of a network and shows another angle of the holistic view that was elaborated decades back. He begins with 18th-century Swiss mathematician, Euler and his "bridges" problem, stopping briefly on "graph theory" (a basis for thinking about today"s "networks"), moves on to Erdos (and Renyi) and their random network theory based on "chance" and "randomness" showing static nature of networks. The link between "Six degrees of separation" by Milgram (experimental psychologist) and its preceding probable origin in Karinthy"s (Hungerian writer) fictional story named "chains" (that first conveyed that at the most five acquaintances connect every person on earth) leads to his own findings: "19 degrees of separation" which estimated Internet to be a small sized world in a way that every document is on average 19 clicks away from any other. Moving from Granovetter"s idea that society is a fragmented web of fully connected "clusters" communicating through "weak ties", to "synchronization" and the quantity "clustering coefficient" by Watts and Strogatz, the author paves a way to his own concepts of "hubs and connectors". The few "connectors" in society are those who know a large number of people like Internet architecture is dominated by a few very highly connected nodes or "hubs". Italian economist Pareto["s] 80/20 (80% citations go to only 38% scientists) rule is related here followed by "POWER LAW" with good examples. An important property is connected to hubs - a consequence of power laws: "scale free networks".

Other properties about networks are pointed through two laws. The "growth" (i.e. networks grow (addition of new nodes), it is compared with Erdos and Renyi"s static and random networks) and the "preferential attachment" (rich get richer: each network starts from small nucleus and expands with addition to new nodes). An answer to: why hubs and power laws emerge in scale free model? And a foundation for theory of evolving networks. Attaching further advancement - "competition" and that each node has certain "fitness", author assigns "fitness" to each Internet node in a network that mimics its ability to compete for links by showing how an early bird is not necessarily a winner (a tussle between "yahoo" to "google"). "Competition" shapes the network topology - it is matched with "Bose-Einstein condensation". Barabasi categorized the networks into two kinds based on topology of networks. The first, in which scale-free topology survives despite the competition for links (fit get rich behavior): coexistence (of hierarchy) of hubs. Examples encompass real networks. Second is star topology - not scale-free (winner takes all the links) with an example of Microsoft. Stress is on scale-free wins most of the networks.

Attention is paid to "robustness of networks" - as in internet, a significant fraction of nodes can be randomly removed from any scale-free network without its breaking apart. Author justifies it with September 11 attacks - why the hubs were targeted? and further explains how the highly infected hubs offer means of persistence and spread based on a large numbers of links connected to it. The examples range from AIDS, to attractive e-mails, to computer virus. Now Paul Baran"s networks (prototype of Internet) find their way with a bit of 1960s-ARPA history. Further, leading to the idea of global internet map, he labels as "success disaster" - a human design that lives a life of its own having underlying fractal structure. The chapter "awakening Internet" concludes with interconnected remarks on growing and evolving Internet at an unparalleled rate paving a possibility towards self-awareness [a parallel to Wolfram"s idea of world being a computer and concept of cellular automata elaborate in his new book "a new kind of science"]. But then where to fit the remotest pockets (say in third world) where social life is still technologically primitive?

In chapter "fragmented web", mapping Internet (size wise) is related with search robots capable of indexing and searching the web with reference to search engines. Two chapters "map of life" (genetics) and "network economy" (economics) are good applications but a reservation is made towards a similar hub on "ecological networks". It is enough convincing that Human Genome Project or Celera Genomics have brought in the "the map of life" but a cellular search engine is still absent. Author shows that ATP is a biggest hub in living system by participating in huge number of reactions. Attention on cancer and manic depression are other analogies here. The last chapter "the web without a spider" emphasizes "no global controller" and shows how a scale-free network is a web without a spider. Once again, author talking about September 11 events and networks to other self-organized instances concludes with the question - where do we go from here? His answer is "un-wrapping", "focusing on dynamics that take place along links and moving beyond the structure and topology towards understanding complexity".

The author"s aim to get you to think networks: how they emerge, look like and evolve is well received. A chapter or big section on already acknowledged networks in "ecological systems" and attention on meta-networks would have added value to this already enough thoughtful, captivating, easy to understand, well designed rather spun web of dynamic efforts and ideas that this book presents. A last touch: the rich "notes" would be better with a separate bibliography.