Customer Reviews

The Social Atom: Why the Rich Get Richer, Cheaters Get Caught, and Your Neighbor Usually Looks Like You

5 star:		(8)
4 star:		(4)
3 star:		(3)
2 star:		(5)
1 star:		(0)

 

 

This book is a wonderfully astute account of recent advances in the social sciences that is beautifully written and accessible to any literate adult. I despair, however, because as a social scientist I have to wonder why it takes a physicist to write such a book. The author might claim that only a physicist could have written this book. After all, the book advances a perspective Buchanan calls "social physics," a perspective that recognizes the free will of individual human atoms but still seeks to explain human social or collective patterns in the manner of physics, "where atomic-level chaos gives way to the clockwork precision of thermodynamics or planetary motion" (xi).

I am exaggerating when I speak of my despair, although I suspect that because many social scientists are needle-nosed specialists, most are incapable of Mr. Buchanan's synoptic vision. I do have a minor quibble with the author though and that is that at several points Mr. Buchanan takes his atomic/collective patterns metaphor a bit too seriously. It is after all just a metaphor. In fact, the brilliance of the book comes through not so much during its description of collective patterns--stock market fluctuations, rumors, neighborhood gentrification, crime waves, ethic violence-- but in the analysis of the features of atoms that make those patterns possible. The analogy with physics is interesting and arresting, but in physics one can literally be indifferent to the properties of individual atoms while explaining collective regularities, but, as Buchanan demonstrates very nicely, human social regularities arise directly from the (universal) properties or propensities of the human atoms. This is a roundabout way of saying that although the book as a whole is brilliant, the most important chapters are chapters 4, 5 and 6--"The Adaptive Atom," "The Imitating Atom," and "The Cooperative Atom." Each of these chapters describes a feature of human nature that is essential to our social lives and helps explain social patterns or regularities.

In the "Adaptive Atom," Buchanan draws mainly from the work of behavioral economists (especially Brian Arthur) that demonstrates the falsity of the "rational choice" model of human beings put forward by neo-classical economists. Rather than the omniscient, logical, calculating automatons neo-classical economists assume we are, the evidence is that we are adaptive agents--we take a step based upon a rule, idea, or belief and then adjust based upon the outcome (63). Our behavior is governed less by deduction than by trial and error. We recognize patterns, make predictions, and then adapt. Our decisions are typically made on the fly.

In the "Imitating Atom" Buchanan draws from a variety of social psychologists, sociologists and neuroscientists, to once again criticize an assumption of neo-classical economists, in this case the assumption that decisions are made by individuals in social vacuums. The evidence is that we are not isolated monads, but, rather, individuals who regularly seek information from others, especially in circumstances of insecurity, ambiguity and danger. Because of this propensity "social cascades" often result: behavior becomes more attractive the more people do it (103).

In the "Cooperative Atom" Buchanan draws chiefly from Peter Richerson, Robert Boyd, Joseph Henrich, Ernst Fehr and Herbert Gintis--all evolutionary thinkers with interests in anthropology and behavioral economics--to drive the final nail in the coffin of "rational choice" theory. The basic point of this chapter is that naturally human beings are not purely self-interested but, rather, "strong reciprocators." We are capable of genuine kindness to those beyond family and friends and we also display righteous indignation toward free riders and those who violate the canons of justice. According to Buchanan and the authors from whom he draws, these features of our nature emerged not through individual competition within groups--such competition favors selfish traits--but via competition between different (cultural) groups. Two decades ago this position was considered heretical among evolutionary thinkers but the evidence has made it a perfectly plausible position among evolutionary social scientists in the last few years and I suspect it will be the consensus view within a few more years among those willing to consult the data.

This brief summary makes it sound as though the book is chiefly an argument against "rational choice" theory, which it is not. Given the importance of "rational choice" theory in economics and, to a lesser degree within sociology, demonstrating the profound failings of the theory is important, but Buchanan also gives "postmodernism" attention, dismissing it as silly claptrap. Additionally, he discusses the tiresome efforts of many social scientists "who have raised the flag of permanent defeat and busy themselves with rehashing the works of great thinkers of the past" (18) and yet other social scientists who mistake the identification of "correlations" for genuine explanations. The most important contribution this book makes, however, is not negative but positive. It is the truly fine summaries of the ideas of thinkers such as Richerson, Boyd, Hernrich, Fehr and Gintis who, along with similarly inclined social scientists, are working to advance and unify the social sciences on a sound empirical basis guided by an evolutionary theory of culture. I recommend this book to anyone interested in the most important recent work being done in the social sciences but I would also recommend that after reading the "Social Atom" one move on to the original sources as well. Richerson and Boyd's Not by Genes Alone: How Culture Transformed Human Evolution or Henrich et al's Foundations of Human Sociality would be especially good places to start.

Brad Lowell Stone

The dust jacket promises that "a groundbreaking social theory reveals the essential - and surprising - simplicity of human behavior" but the book fails to deliver.

What I enjoyed about the book was its ambition to find some underlying concept that would explain much of racial segregation, movements of stock markets, genocide, income inequality and other disparate aspects of society. Buchanan weaves a variety of anecdotes and stories into his presentation that keeps the book from being dry.

My frustration with the book related to both the core theory as well as the format and style. The core theory, as far as I could derive it from his meandering writing style, is that computer programs that model simplified interactions and changes among individuals can be valuable tools to understand how behavior develops. However, most of what he presents is not new or insightful.

Several little things annoyed me:
1) Many of his stories and points have been used more successfully by authors in similar fields. For example, Dawkins has used the tit for tat relationship between the British and German soldiers in WWI's trench warfare in his explanation for the biological basis for cooperation and altruism. Taleb has used the failure of Long Term Capital Management to discuss fat tailed distributions.
2) He creates a number of straw men so that his new theory can knock them down. For example, he repeats several times that until recently economists assumed all individuals were perfectly rational decision makers but I'm hard pressed to recall any economist taking that position in over a decade.
3) An arrogance about physicists pervades the book - there's a clear subtext that people in the financial markets, sociologists and historians aren't as smart as his colleagues. For example, now that a computer model was built in 2007 of how participants in a deregulated Illinois electrical market might work, they now have "real understanding". Now that physicists have looked at US dollar / Japanese Yen exchange rates, their models predict how the rates will move with great accuracy (too many models run by math and physics trained quants have blown up to give this much credence), His attack on the religious on p.200 seemed very out of place in this book.
4) He's sloppy with his facts and conclusions. The Soviet Union did not collapse in 1987 (p.156, try 1991). The collapse of the Russian Ruble did not cause the October 1987 market crash (p.33, blame the Ruble for LTCM's problems a decade later). Although the law of unintended consequences is in full force, his choice of airline deregulation (p.188) and his conclusion that prices are not significantly lower than they would have been and that airplanes are less safe seems plain wrong.

It was my pleasure to read an advance copy of Mark Buchanan's marvelous new book The Social Atom, for which I wrote the following blurb that appears on the back jacket:

"I devoured this book as if it contained the secret answer to the human condition-as indeed it might. To those who have watched the social world unravel in recent decades and wondered why we couldn't do better, Mark Buchanan offers a disarmingly simple solution: emulate the methods of explanation that have already proven themselves effective in the study of nature. The Social Atom is briskly written, informative, and deals with problems of the highest order. Read it and get a glimpse of the coming revolution in the social sciences." Lee McIntyre, author of Dark Ages: The Case for a Science of Human Behavior (MIT Press, 2006)

Now that the book is actually out I'm happy that I can finally say a bit more. One of the best things about this book is that it not only lays out a general philosophical program (be more scientific about the study of human behavior and you'll have a better chance of understanding and fixing a wide range of social problems), it also draws on many specific examples from a number of disciplines to show how this program is already being put into place. One of the previous posters in this forum lamented that it took a physicist to write this book, but in some ways that isn't surprising. Social scientists have sometimes seemed to barely notice the empirical revolution that is taking place right under their noses! This book isn't so much a manifesto, then, as a guidebook to the brilliant work that is already going on in some corners of the study of human behavior, placed within the context of the philosophical foundations that have always supported a more rigorous social science.

Like The Tipping Point and Freakonomics, Buchanan has written a book that is clear and easy to follow and so it will draw the general reader into debates that have been circulating in the academy for years. But he has done so without compromising the depth and complexity of the subject so that, even those of us who have been working on these issues for years, he has advanced the debate. Philosophers, social scientists, the general public (and maybe even theoretical physicists) will learn much from this book.

This is the type of book that I love most of all types. It covers an immensely interesting topic about human behavior, especially about human behavior in aggregate. As a research professor of psychology, I am well acquainted with many of the principles. Further, I believe the topic (if you are interested in it, and likely, even if you are not) is important since we humans are always interested in better understanding why things happen the way they do. Further, it deals with the fundamental issues of group vs. individual, and cooperation vs. competition. (Themes strongly related to my own area of research, which is commitment and sacrifice in marriage and family.) What Buchanan does well is describe a lot of complex phenomena that economists and psychologists have long attempted to quantify in models simple enough to be understandable and workable and complex enough to explain human behavior. The book is strong in laying out an understanding of deficiencies in a rational choice model of human behavior. The book is not as strong in laying out exactly what Buchanan means to say. It is really quite late in the book where you fully can realize what he's really trying to do here, which is to use the spontaneous organization principle that has become popular of late in theories of evolution to explain complex behaviors that appear to have utility and/or design. Further, the book amounts to a current, reasonable summary of ideas about how evolutionary processes can explain something that appears quite incompatible with Darwin: cooperation within groups, and therefore, the survival of groups as well as individuals by selection. Okay, that's what the book is about. It's not clear for a long time that this is what he's trying to do, but he gets there, and it's finally only clear in chapter 6.

Personally, I prefer a little more of a description of the destination before taking the trip, but that is preference. This is why I felt, at times, that the book loses focus and is frustratingly slow in getting to the point. At times a most interesting point is not explained clearly enough even as he's grinding it down from several directions. (Perhaps this is an example of wild fluctuations in the distribution of his thinking which revolves, too often, not tightly enough around part of the central thesis--fat tails. Read the book if you want to understand that point better.) So, what I did not care for about the book is really the author's style. The ideas are interesting. The book excels in the description of various experiments that revolve around explaining things that, at least, appear very complex, such as group cohesion and in-group, out-group behaviors (which are, of course, often quite negative). On the latter point, Buchannan does a nice job in showing how such behaviors serve very positive functions even as they also set up potential for very negative behaviors and acts. The examples are good; the literature cited is often superb. This makes, to me, the book worth the price of admission.

In my own area of commitment, the main ideas of this book have been fruitfully explored by social scientists (especially psychologists) and economists for decades. One clear point of intersection is in the theory or collection of ideas called "exchange theory" which deals fundamentally with the transformation of individual motives and evaluation of gains relative to inputs to cooperative, cohesive, group identity where the gains become associated more with the group's needs and identity than the individuals. My own work in commitment (much of which is translated into a self-help book focused on marriage, The Power of Commitment) focuses heavily on this area, especially noting by theory how transformative it is to have a long-term view on a relationship for all the reasons Buchannan gets at by a different literature in economics. But, it's the same stuff. And it's fascinating stuff. He get's it right that the most powerful transformation seems to come with long-term vs. short term perspectives (along with group identity). All of this applies well to marriages, families, community groups, and on. The hot area in my field at present directly relates to what Buchannan is analyzing, which is sacrifice between partners in relationship. It seems, in many studies in our lab and others to be a particularly salient, positive pattern in romantic relationships. Buchannan focuses a LOT on explaining (or trying to, because it's been a challenge for many for so long across fields) altruism. The emerging idea in my field is that sacrifice may have such particular potency in relationships, and especially marriage and family, because it strongly signals the conditions of non-competition and strong cooperation with the group. Okay, enough of that. Buchannan's book is an attempt to explain by current evolutionary thought what Thibaut and Kelly (1959, 1979) called the "transformation of motivation" or what economists such as Leik and Leik (1977) describes as what happens when one's identity with a group transforms their motivations into non-competition (commitment being, then, the situation between group members who were non-competitive with each other). Buchannan does a nice job of focusing on the other force, the remaining competition forces between my group vs. your group. One could note, btw, that sports have developed as one relatively healthy way to channel what are some of the, otherwise, negative aspects of my groups vs. yours in societies.

The downside to me of Buchannan's book is a tendency to overstatement, hyperbole, and "I've got the secret" that, for me, got annoying for the suggestion of immensely new thoughts, when what he really does pretty well is collect a lot of current and quite old thoughts about how one can explain what, has been considered in evolutionary theory, nearly unexplainable: cooperation and, even, altruism. He occasionally, and unfortunately, does not notice how he does the very thing he's really arguing is bad by the end of the book. The clearest example of this is in the last two pages where he takes some (not unjustified) swipes at religion. They are, of course, not unjustified, but they are unsophisticated. Here is perhaps one of the few places where he sees NO reason (it clearly seems his personal bias) to highlight the positive side of the organizing principles along with the negative. He could, in fact, have done something brilliant right at the end because he could suggest (I am in no way suggesting that he, himself, should do this) that a reasonable scientific endeavor could be to analyze religions of the world along three variables that would, perhaps, be illuminating in further testing his main ideas: tightness of the cohesion forces, rigidity of in-group vs. out-group dynamics, and, despite these forces, degree of emphasis within the belief system on charity to those not in the group. That is good science. If he were aware of much existing science he would find there are already interesting chunks of data here, including the basic fact that, at least in western cultures (where such studies tend to be done), those who are more religious tend to give the most to others who are not even part of their group. It's a pretty interesting question, that goes further steps beyond the conundrum Buchannan wrestles with, to consider how one explains that. It's a worthy question, but by the end of the book, Buchannan seems to show this one strong bias against traditional religious systems of cohesion while failing to note the positive aspects, even as he did for things like forces that lead to genocide. So, it was a bit disappointing that, at the very end, he does the very thing that he seems most motivated (yes, like any scientist, he is motivated toward some message) to suggest is a negative tendency in us all--and especially us humans acting in groups. He fails to his own bias there as well as some lack of awareness of data that exist, favoring the negative clichés as his data. But, all in all, it's a book worth reading provided you are interested in these themes.

"The Social Atom" presents some basic ideas from the field of agent-based behavioral modeling, with examples and applications like stock market pricing, residential segregation, and inequality of personal income. The approach has been around for a while (it got started in the 70s), but now that more work is being done it is a good time for a popular book on the subject.

The strong points of this book are: 1) clear writing for non-scientists, and 2) interesting examples like those mentioned above. However, the book is very simplified, non-quantitative, and lacking in details. Science writing for the popular market is always a trade-off between "too hard" and "too easy," but many readers will want and expect more information than this book provides.

That said, for readers without a science background, this book would be an interesting and understandable introduction to the agent-based way of thinking.

I bought this largely on the strength of the jacket blurb by Mike Davis, which began "Seldom has a book so infuriated me yet kept me tightly gripped to each page." As it turned out, I shared his fury at the author's arrogance, overwriting, meandering, and frequent self-indulgent screeds. As it also turned out, I was not gripped by the book; on the contrary. Though there are certainly nuggets of insight, the book is poorly written and poorly argued. Other reviewers have commented on the free use of straw man arguments and gratuitous digs and on the incoherence of style; I won't repeat their criticisms, though I certainly mean to second them. I barely recognized familiar economic views through the filter of the author's scorn, and found his jejeune comments about post-modernism appalling: no, one does not have to BE a sympathizer with it, but one DOES need to offer a nuanced and sympathetic view of the theoretical perspective one intends to diss. And it is one thing to claim to have found problems with 'rational man' presumptions, and thus to want to emphasize other factors bearing on decisions--as Kahneman so brilliantly does--and quite another to claim that one has identified the SOURCE of these other things...and still another to claim (without argument) that the source is to be found in evolution, as it is understood by the evolutionary psychologists. The wholesale embrace of evolutionary psychology was not defended, nor does the author seem to have any awareness that there are powerful criticisms of both the methodology and substance of evolutionary psychology notably those that claim that it alternates between offering utterly empty hypotheses and utterly ungrounded ones. The weakness of the writing, the weakness of the argument, and the arrogance of the author overshadowed the few, small, good things in the book. As several other reviewers have pointed out, one can find those good things in other places, with clearer presentations and more cogent argumentation.

As a teaching economist, I've often felt that I should read such books as Freakonomics or The Undercover Economist. Occasionally my students have read them and they are surprised, given their commercial success, that I have not. Most students, admittedly, do not do outside reading, but there's also the chance that these authors will give me ideas I can use in class.

The Social Atom is, then, the first of the genre I've read and I'm disappointed. Buchanan rings the changes on old standards such as the Prisoners Dilemma and the Ultimatum Game, but he also allows himself to wander off into geopolitics and religion where his arguments are sometimes unsupported and at other times plain wrong. (Slovenia broke away from Yugoslavia peacefully, and indeed it is quite wrong to say that Yugoslavia itself "had for fifty years been heavily dependent on the Soviet Union... [and under] effective Soviet control" (page 156). This is startlingly wide of the mark, and makes me less willing to trust Buchanan when he strays in this way from his theme.

It's true that, amongst the stories about Tycho Brahe's beer-drinking pet moose and other such trivia, there are several illustrations that were new to me and, in some cases, quite intriguing. I'll note two. The first supposes that each of us has a threshold number in the sense that, if we are in a crowd and a number of people at least as great as this is rioting, then we too will join in. Suppose now that these numbers are distributed evenly amongst 100 people, from 0 to 99. One person will start the riot (having a threshold of zero), another will join in (with a threshold of one), and so on until all 100 people are participating in the riot. However - and this is the point - you need only suppose the absence of that single person whose threshold is one, and the riot never takes hold, never spreading beyond the first person. Buchanan tells this story (from Granovetter) on page 101, and uses it to tear down the representative agent model. On the contrary, he says, "a tiny difference in the character of just one person can have a dramatic effect on the overall group".

If the previous example depends on the idea that individuals imitate one another, the other case I'd like to single out focuses on human beings as adaptive creatures. Taking as his problem the desire we might have to go to a particular bar on nights when it isn't crowded, Buchanan (as always, citing the work of others) supposes that each person starts out with a variety of hypotheses (such as "go to the bar the day after it has been crowded") and, in the light of experience, throws out those which don't work and adopts those which are more successful - an ongoing process, requiring constant adaptation. Elaborating, Buchanan shows that when there are few players, a pattern may persist if none of them happens to have this pattern amongst his or her repertoire of hypotheses; the pattern will go unrecognized. When the number of players increases, though, the pattern will be jumped on by someone, whose success will in turn lead others to adopt it. With few players, then, there is a pattern; with more, there may be only random change. I must admit I found this illustration (on pages 74-5 and 83) to be well-explained and intriguing.

Overall, I'd suggest that people like me, who could do with a brush-up on their experimental social psychology and experimental economics, look for a book which develops examples such as the ones I've mentioned, but in a more focused way. And as for my students, I can't recommend this book to them since, in spite of its readability, I don't find it to convey very much that is either new or fundamental.

Much like the 80/20 rule that came from Italian economist Vilfredo Pareto which was created to describe the unequal distribution of wealth in his country, the Social Atom expands this to the social sciences on broader terms. The book is a very quick read from both a Psychology and Social Science perspective, but also on the hidden patterns and explanations of why things appear as they do.

Each chapter essentially builds on itself to get the reader to the point where the author can then discuss the "critical state" of things. One example is you can randomly place yellow, brown, black, and white objects in an environment with very simple rules and the objects will, over time, tend to join each other of similar color. Now replace the objects with people and you get the same clustering in most every major city which has nothing to do with racism.

As most ecologists now know, if you destroy one pest in farming, you may introduce another more menacing pest going forward or the unintended consequences of a good idea. The Social Atom helps the reader understand that small steps/changes can cause a phase change, like ice to water. These phase changes indicate that things are indeed in the critical state and subsequently obey the results of power laws like Vilfredo Pareto's wealth distribution which again is applicable to almost every nation whether democratic or otherwise of a different ideology.

The point is that many hidden patterns and systems inherently evolve to the critical state and if you look closely enough they are present everywhere (i.e. Ubiquity), but we just never noticed before. See Ubiquity: Why Catastrophes Happen

This is a great easy to read book looking at human dynamics from a complexity perspective. Often we need to look at underlying patterns common to all systems rather than delving more deeply into an individual's psychology or whatever. The book links the physics of the atomic structure and compares it to human dynamics. It introduces power laws, game theory, fat tail graphs, Pareto's law, fractals, etc. It is a great first book to read.

I was surprised by the comments at the very end about religion. From nowhere there is a discussion about religion as a different way of making sense of the world that has resulted in mass slaughter. He quotes Sam Harris. There is no discussion of any spiritual understandings other than traditional religion.

Buzz Holling's adaptive cycle and Spiral Dynamics could have extended the theme of the book to have been more coherent and explanatory of the ideas

I read the Social Atom by Mark Buchanan after following a reference to it from the Financial Ascent of Money by Niall Ferguson. As a scientist (chemist) myself, I was impressed by his ideas about how individuals and societies are evolutionarily hard-wired to behave like they do. However, I found his references to 'scientific' phenomena mostly a stretch. Even in cases when he does hit the nail, he does not hammer it enough to explain the phenomenon enough to appeal to people with advanced science degrees. As a former physics editor of Nature, I expected he would delve into the physics aspects, particularly quantum theory to draw parallels with our world. For instance, the same wave function that defines electron movement around the nucleus of an atom also explains tidal phenomena. the author also tends to belabor points constantly throughout the book. I would have preferred more sharp, concise comparisons and letting the reader do the rest of the thinking.

Overall this book is a good read. And small enough to finish over a short haul flight. The particular point that stuck with me after reading the book was that great politicians and statesmen are not usually the best and brightest amongst us, but the individuals who have been able to recognize and exploit social patterns. Recommended!