Customer Reviews

Bursts: The Hidden Pattern Behind Everything We Do

5 star:		(5)
4 star:		(8)
3 star:		(8)
2 star:		(16)
1 star:		(9)

 

 

I thought Albert-László Barabási's first book, "Linked: How Everything Is Connected to Everything Else and What It Means," was excellent (see my 4/18/10 review), so I looked forward to reading "Bursts" with great anticipation, hoping that he was going update us on all the interesting things he learned in the intervening 8 years (especially related to biomedicine and cancer). Instead, having just finished "Bursts," it's hard to convey how disappointed I am.

While "Linked" presented plenty of solid and useful science in an appealing format, "Bursts" has minimal scientific content and I learned almost nothing. The only significant idea Barabási presents is that the time-spacing of many events in the natural and artifical worlds follows a power law distribution, which means that events have some tendency to cluster into "bursts," although very widely spaced events can also occur, since power laws have "long tails" rather than dropping off exponentially (as Barabási himself acknowledges in passing, "bursts" is a somewhat misleading term, since power law distributions are continuous, not dichotomous). But Barabási doesn't offer much explanation for the ubiquity of these power laws, nor does he offer useful insights regarding their implications.

He does try to argue that awareness of these power laws will eventually enable precise prediction of human behavior, but this is simultaneously both obvious and wrong (and it's telling that Barabási appears to be unaware of the seminal work of Quetelet on this topic). It's obvious because we already know that people are necessarily creatures of routine and habit, so where we are and what we're doing will often be predictable. But it's wrong because, like the weather, our lives also involve volatility and bifurcation points, such that much that's important about our individual and collective lives will remain unpredictable. I've experienced this in my own life in profound ways, and so have you (think back, and you'll recall some pivotal moments).

Most of the book is actually taken up by a discussion of an episode from Hungarian history of the 1500s. This may interest Barabási for personal reasons, and perhaps it satisfies some urge to be a historian or novelist (which he apparently has a knack for), but it has no place in this book. I kept waiting for this plot and other plots interwoven throughout the book to all gel together in the end, but they never did -- I feel like I was waiting for Godot.

Overall, this book was a waste of money and (more importantly) time. The only redeeming feature is that I was able to read it quickly (three days), but that's small consolation. I really don't know what Barabási was thinking. I must also add that I was partly swayed to read this book by the endorsement from Nassim Nicholas Taleb on the back cover; that endorsement has unfortunately harmed Taleb's credibility in my eyes.

I'm not the only reviewer expressing strong disappointment with this book. Barabasi did such a masterful job with Linked that I grabbed this book to moment I saw it hoping for more of the same "can't put it down" reading I got with Linked. Wow was I surprised--and not in a good way.

This book is not Linked nor anything like it, beware! Linked was a very well crafted story that explained various complex topics about networks. Bursts is pop-science at best. Worse, like other reviewers, I too am totally annoyed by the Hungarian history lesson from the 1500s that takes up every other chapter. And, after reading one of these POINTLESS chapters you feel like, "what was this supposed to add to the overall understanding of the topic?" The answer is little to nothing.

Bursts lacks the insightful and useful science that Linked gave us. The references/notes are OK, not fantastic like in Linked. The book uses a lot of text to make some rather simple points about behavior. We behave in bursts, not randomly. Bursts (activity clusters in time) exhibit power law characteristics. In the future, since behavior is not random, perhaps it can be better predicted.

Save your money and your time! At best, wait for the paperback (if it makes it that far) and read it at the beach. Better still, do a second reading of Linked, you'll get more out of it that you would Bursts. Linked: How Everything Is Connected to Everything Else and What It Means

Here in the Bronx there is a street scam called three-card monte. A person skillfully manipulates three cards, moving them around on a makeshift table, usually the bottom of a large, upside down cardboard box. The object of the game is to pick the ace among the three cards, after the scammer comes to a full stop and lays the three cards face down. To lure an innocent victim, the shuffler has two or three partners stand around the box, pretending to play. One of the partners will then "guess" the right card, and the shuffler will "pay" him or her $20 for the win. Naive innocent onlookers will then play and lose their bet as the skilled shuffler will do whatever trick it is he does to ensure that the victim does not win. All along during the shuffling the victims are flashed for an instant a view of the ace, which then disappears, never to resurface after the victim makes his or her choice.

Reading Bursts felt something like being taken at a three card monte game. The author jumps back and forth between a convoluted, though admittedly interesting, historical epoch in Hungary, and then back to studies and analysis done about different behavioral phenomena, like the way people use their cell phones, or respond to email or correspondence, or how dollar bills circulate, and then back to the Hungarian episode, and then back to the research, and now we hop back to a story about a Muslim surnamed individual who seems to be an exception to the author's findings, back to the history, back to more research, and on and on. The ace among the cards is Barabasi's claim that people tend to behave in "bursty" ways, that is doing some things intensively over a short period of time, and then doing nothing or very little of that thing for a long time. I don't particularly see why this is so revolutionary.

Knowing that we behave like this will supposedly, at some time in the future when the technological stars align properly, allow prediction of human behavior and we will enter the realm of the creepy, where our future whereabouts and actions could be tracked for nefarious purposes. Perhaps this is true, and if so, significant, but Barbarasi expresses very weak conviction about this, consistent with his waffling and vagueness about other details of his "breakthrough" discovery. Because of all the above, the reader is left with the feeling that someone has tried to elevate commons sense to some esoteric art and perhaps has been duped by three card monte social science.

Barabasi is a well-known researcher whose scientific work on human networks is highly regarded. In this book he uses his mathematical work as a takeoff point to demonstrate that with enough observations one can "predict" the whereabouts of anyone. The argument is unconvincing and the examples are contrived and often incoherent (e.g., a peripatetic artist friend whose life seems to consist almost entirely of close encounters with Homeland Security). As for predicting one's future from numerous observations of one's past, another reviewer has pointed out that the Belgian polymath Adolph Quetelet tried this a century ago, gathering immense quantities of anthropometric and sociological statistics that he thought, among other things, could predict criminal behavior. On a population basis this almost works because all sorts of environmental variables are correlated with each other and the great majority of people in the world lead lives of humdrum routine, but it fails on the individual level, in part because it utterly ignores the unique social and physical environment that every individual inhabits. Stale and vacuous stereotypes abound, such as the "prediction" that MIT students can be found in their dorms on weeknights and in bars on weekends. When I went there, on any given night students could be found at fencing practice, rehearsing with the MIT Symphony, editing The Tech, or sailing on the Charles. Near the end of the book Barabasi makes the fatal error that many better writers before him have made (e.g., Jeremy Rifkin) by trying to anthropomorphize concepts of statistical thermodynamics, especially entropy and the Heisenberg uncertainty principle. This technobabble may easily impress people without a doctorate in physical chemistry, but to someone who actually knows the field Barabasi's pronouncements are gibberish. His prediction that depression will soon be the leading cause of death demonstrates breathtaking ignorance of the most elementary concepts of both mental and physical health. Then there's his foray into Hungarian history, a pointless distraction that takes up half the book. Overall, the book is disorganized and relies far too much on the overbaked formula of "golly-gee-whiz all the experts were wrong."

The author was born in Transylvania and would like us to know all about the 16th century battle of Koloszvar, where his "heroic" ancestor, Lenard Barlabasi, led a bunch of well-armed knights in a massacre of mostly poor peasants who were trying to bring a new order to greater Hungary. Since only about 3 people in the world care about this obscure corner of Transylvanian history, Barabasi has to come up with some excuse for a theme that would find readers for his genealogical excursion, so he recycles the 30-year old issue attention cycle to pretend he has some new theory about "the hidden pattern behind everything we do." The excuse to bring in Transylvania is a pronouncement by another Hungarian councilor, Istvan Telegdi, who supposedly predicted the whole sequence of events that led to the Koloszvar battle beforehand. Barabasi pretends to want to know how it is that Telegdi could have predicted the events that led to Koloszvar when supposedly humans behave randomly. It turns out that a newly elected and power-hungry de Medici Pope, Leo X, wanted to keep a Hungarian papal challenger at bay and came up with the idea of a peasant Crusade to retake Constantinople in order to have an excuse to get the challenger out of Rome, leading to a whole cascade of events that ended up at Koloszvar with Hungarians slaughtering one another. Of course recruiting peasants in April for a Crusade before the harvest would not sit well with the knights they worked for, and encouraging peasants to fight instead of the knights whose whole excuse for a privileged societal position was their supposed availability for Crusades like this one was a recipe for disaster. The answer to how Telegdi foretold what would occur is two-fold: #1, it is obvious, and #2, there's a good chance that the whole story about Telegdi's prophecy was made up later. In other words, nothing about this whole lengthy discursus, which takes up a good two-thirds of the book, has anything to do with the title. The rest of the book consists of random musings about, e.g., the incompetent/evil U.S. government that, post-9/11, keeps picking up for questioning a peripatetic American with a Muslim name who insists on travelling to every terrorist capital in the world and who Barabasi recognizes is an outlier in his travel patterns. Also a bunch of random thoughts about the fact that we send email and view web pages in bursts (well, duh). Unless you have some specific interest in 16th century Transylvanian history, stay far, far away.

This book was a huge disappointment based on my expectations from Barabasi's prior book, Linked. The scientific elements of this book were piddling. The multiple stories told piecemeal throughout each chapter were akin to listening to someone with ADD tell them. None of the stories appeared to affirm or even correspond to the few scientific explanations within the book. In the end (literally), the only memorable chapter was the horrid description of the execution of one of the main characters whose tale was told in fits and starts throughout the book. I only kept reading this book, because I kept expecting, based on Barabasi's earlier book, Linked, and my great respect for him as a scientist, that things would be neatly tied together eventually. Unfortunately, they weren't. This was a book that overpromised and under-delivered.

As a child, I devoured cold cereal to get to the prize at the bottom. The cereal had no nutritional value, and the prize never met my expectations. Bursts is that cereal, though the little prizes are better than expected. If a surprise glimpse into Hungarian history is for you, it might be worth reading the rest. But if you are interested in the science, forget it.

Barabasi does not give a coherent story of the predictability of human activity. Instead we are treated to episodic descriptions of findings, contradictory findings, and findings contradictory of the contradictory findings. Human actions occur in bursts -- unless they do not -- and follow power laws -- except when they do not -- and can be modeled by Levy flights -- though perhaps not. We do not get to see how the studies were conducted, nor do we get a view of the interior logic theories. We do learn about the social life of the author. And just when he introduces an interesting topic, rather than delve into it, he veers off into Hungarian history or the travails of an artist beset by Homeland Security.

So what is predictable about humans? If a book has a one syllable action noun for the main title, then there is a very high probability that the contents are mostly sugar and filler. What is less predictable is the small plastic prize at the bottom.

Sorry, but this book is not worth to read. Long parts with pseudo-pop-science gibberish combined with chapters of Hungarian history makes this book incomparable to great "Linked", which I would recommend to anyone. There are (in Burts) some references to interesting researches and some good (small) points, but thats all.

The book seems to be a collection of unrelated stories of history. I finished it waiting for what the title promised. Only one chapter on Power Rules (ch 21 - yes, you read that right ... Chapter 21) has anything to do with the topic or Bursts.

Yes, you need to slog through 20 boring, incoherent chapters from the 1500's to get to the meat. And after chapter 21 ... more non-sense.

Really, save your money ... I want mine back!

Having read and enjoyed Linked very much, I was eager to get my hands on this book. The book was easy to read and maintained a level of tension throughout, as you might expect from an action series type of fiction. While Linked piqued one's interest in Power Laws and Networks and gave us ample leads to pursue these, Bursts tended to stop you from taking the leads! The Hungarian folklore did not do much for me other than that it was a good story. As I read the book I got the feeling that the author was setting up theories only to be proven wrong. From the point of view of a scientific pursuit, it is acceptable to define, test, validate, retest, and invalidate (under new evidence) new theories. Of course, at the end of the book I was left wondering if any of the theories are valid. Do we have Power Law, Einsteinian Random events, Poisson distributions, Levy plots, bursts, or what? If the author's intent was to make the reader ponder these then he achieved his goal.

By the way, though the review says it is a Kindle book review, it is a paper book review.