Forgot your password?
typodupeerror
Science Books Media Book Reviews

Emergence 103

Posted by timothy
from the good-heavy-light-reading dept.
mrgrumpy contributes this review of a book which, despite the interesting subject matter, he says comes with some forgiveable but hard-to-ignore shortcomings. Emergence still sounds like an interesting read and broad introduction to the theories and subject matter at hand, but read on for more of his take on the book.
Emergence - The connected lives of ants, brains, cities and software
author Steve Johnson
pages 288
publisher Penguin
rating 6
reviewer Jim Richards
ISBN 0-713-99400-2
summary Seeing order and patterns in apparent chaos.

This book covers the theory of emergence, which states that within a system of what seems to be anarchy, there are underlying rules that govern the pattern of behaviour and bring order out of chaos.

This books serves as an introduction to the field of emergence. It is something that is already happening around us, but we usually cannot see. The reason for this is that you need to look at a higher level then the individual organism. Ants can not see the society as a whole that they are members of. Just as we humans may have an understanding of the local community we are in and of ourselves, we need to step outside (or above) the city to understand how it functions. A city, like an ant colony does not have rules from the top as such, but rules that each occupant obeys, and it is these rules that give order to the chaos and make the resultant community behave like an organism as a whole.

I really wanted to like this book. But the level of information within it will make me put in into the light, popular fiction section of my bookshelf. One of the aspects of the book that really wanted me to give a good review is that the author makes a good introduction to the theory behind the comments system of Slashdot, the way people are chosen to rate comments and how good comments filter to the top. As such, I would have liked a review of the editorial process on Kuro5hin as well, since the two systems as fairly similar. In fact, I think the Kuro5hin system is better, because long time readers will see that the stories have moved away from an open source/linux focus to more cultural aspects, thus reflecting the change and growth of the community. But the idea of a Daily Me portal, that serves information that would suit us is explored heavily.

As I read the book though, an uneasiness came upon me, just as I do when reading books on neo-Darwinism. There is no mention of where these rules as such come from except through evolutionary survival or initial chance. If anything, the author implies that we are in a universe that had the initial conditions set, and left running. So we'd evolve or grow into who or what we are.

The idea that a God figure could be there, tweaking the parameters as the model runs, or even setting the initial conditions works against his ideas. This view is however explored in the chapter Control Artist, where the author comments on the development of software models, notably computer games. Games such as SimCity are discussed where the rules are set, but as a player we get to choose what gets built, what gets destroyed. Although here we are playing the Mayor of the City, the notion is the same; we control the macro level and not the micro level. But at the micro level, the software developer who built the game in the first place controls each inhabitant. Nothing really, is left to chance. Given the exact same initial conditions and same set of instructions the computer will create the same environment.

So, like most popular science books currently available it will educate you, entertain you and keep you occupied while reading it or totally bore you. But it is not a book of philosophy to base life on, which thankfully, the author has not tried to provide. It is very well researched, and the author seems on top of current trends and ideas. His writing style jumps around quite a bit, and some of the connections between topics might seem a little far fetched but it is an entertaining read as an introduction to the field of emergence theory.

Pet peeve 1: Notes. The notes section at the end is fairly extensive. But there are no foot notes in the book. The notes are indexed by page and quote. So as a reader you have to constantly check the notes section to see if there is a note or reference for the page you are reading.

Pet peeve 2: There was (for me) a glaring technical error on page 120.

"Ironically, it is precisely this feedback that the Web lacks, because HTML-based links are one-directional. You can point to ten other sites from your home page, but there's no way for those pages to know that you're pointing to them, short of you taking the time to fire off an e-mail to their respective webmasters."
You can see who is visiting your site, unless they are using an anonymizer proxy, or other system to hide your headers. The HTTP-REFERER header gives you exactly this information.


You can purchase Emergence at Fatbrain. Want to see your own review here? Read the book review guidelines first, then use the web submissions form.

This discussion has been archived. No new comments can be posted.

Emergence

Comments Filter:
  • by flimflam (21332) on Thursday December 13, 2001 @11:30AM (#2699213) Homepage
    not that the links actually exist. A link is, in fact, one directional. If no one follows a link, there is no way to know that it exists. Practically, looking at the referrers (or should I say 'referers' to use the official but wrong spelling) of your HTTP requests will tell you pretty much the same thing, but there's a conceptual difference between that and actually having some sort of "reverse-link". Kind of like asking everyone who comes into your store where they heard about you as opposed to hearing first hand from the people making the recommendations.
    • If no one follows a link, there is no way to know that it exists.

      Well, there's always google: try a query for link:your.link.here. It will return a list of pages that link to your link. It's not inherent to HTML, but it's a way, and without need for referers.

    • Referrer tells you who's following the links not that the links actually exist.

      Bzzt! The referer header tells you the URL of the document containing the link. It tells you nothing about who followed the link. A referer header can be forged, but otherwise there actually is a link and it is contained in the document whose URL is in the referer header.

      • You can see the IP just fine in the logs, so you do know who has followed the link, though not by the referrer field, you're right there.
        But the point is that you never know the links that exist, only the ones that are followed. In this sense linking is one-directional.

        Searching google for links is another matter, but it's not a "glaring error" to claim that linking is one-directional. Debatable is a more appropriate word.
      • Bzzt! The referer header tells you the URL of the document containing the link. It tells you nothing about who followed the link.

        I think you misunderstood the comment. He meant that you only receive information on those links that people follow. A hundred pages could link to your site, but if nobody ever follows those links, you'll never know they exist.

        The subject of his message may have been confusing. But if you read the rest of you'll see that he didn't imply that HTTP-REFERER gives you additional information about the person who followed the link, it only provides the URL of the page they came from.

    • HTML links which use the A element are unidirectional primarily because they are embedded in the source document (the webpage at the start of the link).

      There is a branch of hypermedia research which deals with what is known as Open Hypermedia, in which links are objects which exist independantly of the documents that they link. This allows more complex link types, such as bidirectional or n-ary (many-ended) links, and promotes a degree of flexibility in the hypertext because links may be applied to many different documents (simplifying link maintenance).

      Because the links are stored separately from documents, the resolution of links can be performed both forwards (when following a link in the normal manner) and backwards (when asking "what links point at this page?") with equal ease. The relative paucity of the Web with respect to this type of links has historically been an issue in the hypermedia research community, and it is due to this that the W3C has taken steps to rectify the situation by introducing the XLink [w3.org] recommendation, which allows the creation of open hypertext links.

  • You can see who is visiting your site, unless they are using an anonymizer proxy, or other system to hide your headers.

    If no one clicks on the link to get to the page, you will never know the link exists. Do those links matter? Probably not.
    • Somewhat OT, but google.com come's to the rescue (as I'm sure most of the big search engines do) with their "Who's linking to who" searches. To see who's linking to slashdot, for example, try... http://www.google.com/search?as_lq=slashdot.org&bt nG=Seerch [google.com]
      • And that's the point that everyone is missing. HTML links are not inherently two-way. It takes some extra mechanism: your browser voluntarilty telling a site from where you are finding it, or a monster like Google scouring everything to look for who links to whom.

        The point being made in the book, I believe, is about Google. It uses the phenomenon of linking in order to rank sites. There is no "master thinking program" analyzing page contents to determine similarities. Rather, each page's links (and the text they use to link) come together (in Google) to "self-select" similarities. That a larger taxonomical "intelligence" emerges from the individual phenomenon of linking is evident from how damn right Google gets its results.
    • [sarcasm]
      How else can you sue people for accessing your content in 'inappropriate' ways?

      If someone came to a deep page without first going through the 'EULA' page, or some other crank..
      [/sarcasm]
    • If a link "error 500"'s in the woods, with nobody around, does anybody hear the sys admin scream?
    • Well, since some browsers let you right click on a link and add it to your favorites, I frequently see a link that I don't want to follow now, so I put it in a folder to visit later. That means I found out about PageB from PageA, but there's no way for PageB to see that PageA. I also copy and paste url's.
    • Yes they matter very much to search engines like google which will rank your page according to the number of other pages pointing to it. It does not necessarily needs to follow it.

      Hey, this is my first /. post; I need a sig !
    • If a page linked to in the world wide web, but nobody clicks on it...Is it still a web page?
  • You can see who is visiting your site, unless they are using an anonymizer proxy, or other system to hide your headers. The HTTP-REFERRER header gives you exactly this information.

    Technically this is not true. The HTTP-REFERRER only shows you who sends people to your site by a link, not who links to it. Just because you can tell where someone came from doesn't change the structure of a link. The link its self is truly one way.

  • I've read materials somewhat akin to this in the past, and they all seem to boil down to the "if you stand back far enough, it looks like order..." argument. But they always seem to overlook that inescapable fact of probability - if you look at ANYTHING on a large enough scale, you'll begin to see some order. Kind of like the "infinite monkeys/infinite typewriters" adage. That is, if you put an infinite number of monkeys in front of an infinite number of typewriters, at least one of those monkeys is going to type the complete works of Shakespeare. Sounds stupid at first, but in reality, it's absolutely true. Infinite scope leads to infinite odds in favor of what you're looking for (or not looking for).

    By the way, did anyone else notice that this review sounded more like a school book report than an actual review? The guy who submitted spent an awful lot of time pointing out stupid "technical errors" (they weren't errors, by the way) to really lend any plausibility to his review. Just a thought.
    • by pmcneill (146350) on Thursday December 13, 2001 @11:48AM (#2699300)
      Emergence, which the reviewer didn't seem to cover very well, is not "keep moving backwards until you see something ordered." It is a behavior (usually, from what I've read about, in groups) where rules that pertain only to individuals give rise to complex behaviors.

      The canonical example of this is flocking, from Flocks, Herds, and Schools [toronto.edu] by Craig Reynolds. Basically, if all members of a group avoid collisions (with obstacles and each other), match velocity with locally perceived group members, and stay close to local members, a flocking behavior is achieved. There is good evidence as well that this is how flocking is achieved in nature -- ornithologists (bird people.. may have gotten the word wrong:)) have studied Mr. Reynolds' simulations and found them to be indistinguishable from birds.

      Another example is ants finding their way to food. They start off randomly travelling, leaving their pheremone trails everywhere. However, once food is found, the pheremone trail to that food is reinforced over and over, causing more and more ants to travel there. Eventually you see a line of ants going straight for the food -- all because of a "wander unless I sense enough pheremones" behavior.

      The point of this is that emergence is not necessarily a global phenomenon -- it occurs at all levels.
      • Sorry but I do not see ants following a chemical trail as being a good example of emergent behavior. What is the higher-level activity resulting from the collective behavior of the lower-level activity? Now Aunt Hillary... she's one emergent Aunt!
        • The higher level behavior is "feeding the nest." The ants are bringing the food back, so the pheremone trail is benefiting the nest as a whole. The trail connects the nest to the food.

          -- Your body is just a flock of cells flying in close formation.
          • Thanks for your kind reply. I still do not see that this is emergent behavior in the same vein as, say, mind or consciousness might be as a result of the cellular activity of the human brain...although I happen to believe that the process occurs the other way around - (matter as a projection of consciousness). The notion of flocking birds is interesting. I understand that the most reasonable way to simulate this activity with a computer program is to implement it something like the way Conway's game of life is implemented. Probably the same with simulating swarms of fish (getting my money's worth cable). These two phenomonon display group qualities (emergent behavior) that are visually stunning but not as interesting (to me) functionally. Perhaps it is a matter of degree.
        • The emergent behavior is that if a single ant finds a food source, the entire hive will be alerted to the fact and find a very short route to the food, all using only simple chemical markers.
      • The problem with this explanation is that it is, at best, incomplete. If the only things members of a flock do is avoid collisions, match velocities, and stay close to perceived group members, why is it that flocks actually go somewhere rather than merely fly around randomly?
        • These rules don't take into account the initial conditions and the motivation of individuals. Consider the situation where birds are sitting on a wire and one decides to fly off. Others ready to leave might join a flock with that bird -- it already has a velocity to match, giving it motion. Another situation is the lead birds getting tired. They may head towards a tree, and the rest of the flock will follow. You're entirely correct that if the intial condition is just a bunch of motionless birds suspended in air, the flock very likely will just swarm around that one spot.
          • I think if the initial conditions were a bunch of motionless birds suspended in the air then the next thing to happend would be synchronised plummet as they wonder who the fuck was doing thought experiments on them. I guess they would eventually swarm around a lower point, unless they impacted first.

            Just ignore me.

      • Emergence, which the reviewer didn't seem to cover very well, is not "keep moving backwards until you see something ordered." It is a behavior (usually, from what I've read about, in groups) where rules that pertain only to individuals give rise to complex behaviors.

        I'm not saying that Emergence is that. I'm saying that that's what the book identifies as being the way to observe it.

        You haven't read the book, but your idea about emergence is only half way to what the author is trying to identify. It is about the behaviour of individuals within a community, and how that community appears. The emergence behaiour can only be seen at a macro-level, not at the individual member level.

    • by Tackhead (54550) on Thursday December 13, 2001 @12:01PM (#2699379)
      > Kind of like the "infinite monkeys/infinite typewriters" adage.

      20 years of USENET archived, and we still ain't seen it.

    • if you look at ANYTHING on a large enough scale, you'll begin to see some order.

      Good point!

      A couple of years ago I was fortunate enough to attend a lecture series on emergence by Prof. Benoit Mandelbrot. When he talked about clusters of galaxies, he posed the problem: 'Are the patterns out there, or are the patterns only in our minds?' In nature without conscious observers, there are no patterns - because there is nobody to define clustering or other patterns. And if we define 'order' in a different way, we start seeing different things.

      It seems that the human vision is inherently 'tuned' to see certain kinds of order in everything, and it has probably been useful over the course of evolution. There are several stories of people who have seen the face of Christ on a randomly colored surface. Sometimes when meditating, I stare at the carpet and see weird forms emerging. In the million-monkey experiment, you probably start to see interesting stories way before the emergence of a grammar-proofed Hamlet.

    • But they always seem to overlook that inescapable fact of probability - if you look at ANYTHING on a large enough scale, you'll begin to see some order.

      But that's not really what emergence is. Emergence is large scale behavior that is not built into the small scale rules. Traffic, for instance, is an emergent behavior. There is nothing in the descrption of a single car driving around that makes traffic necessary. However, certain conditions will invariably lead to it's development.

      Emergence is nested all the way down, in every physical system. Society is emergent from individual behavior. Individual behavior is emergent from biology. Biology is emergent from chemistry. Chemistry is emergent from physics. Physics may be emergent from the fundamental law of the universe. I might just have to buy this book. :)
    • if you put an infinite number of monkeys in front of an infinite number of typewriters, at least one of those monkeys is going to type the complete works of Shakespeare
      Despite the popularity of this myth it has a substantial problem in that the number of possible keystroke sequences grows incomparably faster than the number of monkeys, wordprocessors (surely we wouldn't rely on typewriters in this day and age) and keystrokes. So no matter how many monkeys and typewriters you have, don't expect to produce more than the odd short string which matches exactly the words of the Bard.
      if you look at ANYTHING on a large enough scale, you'll begin to see some order
      Whatever order you see appears for other reasons. Ignoring the ability of our perception processes to provide an illusion of order in response to initially random stimuli, as in dreams, it is population-wide phenomena like emergence and evolution which produce large scale organisation. These phenomena are actually much more universal than the laws of physics, applying analogously in physical, biological and social worlds ... even in computer simulations.
  • The idea that a God figure could be there, tweaking the parameters as the model runs, or even setting the initial conditions works against his ideas.

    You can hardly blame the author for writing a popular science text that fails to include wild speculation about the influence of medieval superstition on physical phenomena. I imagine he probably failed to consider phlogiston and the luminiferous ether, too, but I would have a hard time holding that against him either.

    • I believe that what the review writer is trying to say is that the author _did_ imply that a God figure may be there setting initial conditions, and the fact that the author implied this took credence away from otherwise good ideas.
  • by El_Che (161286)

    Er, interesting review.

    I googled and filtered, an intro to Emergence the notion, [mit.edu] and an excerpt from Emergence the book. [nynma.org] (In which Slashdot is discussed.)

    Oh, and here's a less interesting book review of Emergence from the Village Voice. [villagevoice.com]

  • by Chocky2 (99588) <c@llum.org> on Thursday December 13, 2001 @11:48AM (#2699299)
    Like the reviewer I wanted to like this, and it does manage to serve as a passable overview of evolution in complex systems, but only in the same way as "A Brief History of Time" serves as a passable introduction to cosmology.

    The primary problem is that rather than being a popular science book, it comes accross as a populist one, picking easy pop-culture references rather than more appropriate ones. This would be more forgivable if the book was giving a more balanced view of the subject, but the author seems to have a definate agenda which he is trying to communicate rather than giving a solid, unbiased view of the topic.

    A good book on the subject for people who have no knowledge of, or interest in, the topic.
  • Steven Johnson? this is the same guy responsible for Interface Culture: How New Technology Transforms the Way We Create and Communicate [fatbrain.com], a book i've had the misfortune to read. Mr Johnson seems to have the Wired Magazine-style flair for making mountains out of molehills, making blind reaches into that which he knows nothing about (i.e. physics), and making an ass out of himself. do yourself a favor and read something smarter (maybe something from the Oprah Book Club?).

    (the fact that most of the comments are clarifications on the HTTP-REFERRER discussion seems to suggest that the book might not terribly engaging to the Slashdot audience anyhow.)
  • Not only are links one way with HTML (excepting the http-referrer info, and this can be forged) but in fact for most site-owners, they themselves don't even know where they are sending people. Without using some database-backed link-tracking bouncer, you as a website designer have no idea who is clicking on what links. The times that this is most useful is for advertising because you are most likely getting paid per-click. Not only is it beyond the scope of most basic homepage authors to do this (or impossible depending on where you host), but also you then have a single point of failure -- when was the last time you clicked on a banner ad link or similiar just to have the script be broken in some way and not actually 403 you to the right site.
  • by DaoudaW (533025) on Thursday December 13, 2001 @11:53AM (#2699329)
    Since reading and doing an undergrad seminar on Symmetry in Chaos by Marty Golubitsky [uh.edu] and Mike Field [uh.edu] several years ago, I've been quite interested in this topic.

    A more serious alternative to Emergence might be The Self-Made Tapestry: Pattern Formation in Nature [santafe.edu].
  • by JMZero (449047)
    This idea of emergence seems different than the one I always imagine.

    I think of emergence in terms of complex behavior resulting from simple rules (eg. the many kinds of human thought resulting from the interactions of a pile of simple neurons).

    I think of emergence in terms of "the whole is greater than its parts" rather than "there's order in chaos".
    • I think of emergence in terms of complex behavior resulting from simple rules

      This is a nice way of thinking emergence, and a very useful one even in plain old classical physics. For example, the macroscopic properties of matter, such as temperature, are emergent in the sense that a lower level description of the system can be discarded. Temperature is not really a property of any of the system's parts, but one of the system as a whole. Temperature, pressure and such are a result of the collective behaviour of the molecules, and in systems where the interaction between the parts (molecules) is significant, the resulting behaviour is complex.

      I think of emergence in terms of "the whole is greater than its parts"

      I don't really like this phrase, or the variant "the whole is greater than the sum of its parts". The one I mention is more evil because it uses a very specific mathematical concept "sum" in a very vague manner. But it is clear that if I take the parts of a car, for example, and stuff them into a box, they're not a car. They are the parts of a car. I don't see what's the deep philosophical insight in a phrase like this. It's just stated in a mysterious way to conceal the fact that nothing has really been said.

      On the other hand, "The whole is equal to its parts and their interactions" sounds a lot less flashy.
  • by richieb (3277) <richieb&gmail,com> on Thursday December 13, 2001 @11:56AM (#2699346) Homepage Journal
    I read this book few weeks ago and found it pretty shallow, not much info, although few interesting ideas. If you are looking for more depth I recomend: Turles, Termites and Traffic Jams by Mitchel Resnick. Resnick is a professor at MIT doing research on these things and is the author of StarLogo - a massively parallel version of LOGO.

    • A massively parallel version of LOGO? Presumably to control your massive army of one-axis robots?!?
    • I just finished reading Resnick's book, and I'll give it a hearty second recommendation. It is an excellent (and fun!) introduction to decentralized thinking. It's very lucidly written, short, yet conveys its core ideas very well.

      As the prior poster mentioned, StarLogo is a "massively parallel logo" -- a language inspired by Logo and retains Logo's coneptual simplicity. The goal of StarLogo's design was to create a tool in which to explore emergent behavior and decentralized thinking, yet remain a very accessible and fun environment.
  • Emergence (Score:3, Insightful)

    by pogofish (514289) on Thursday December 13, 2001 @12:00PM (#2699370) Homepage

    I've read several popular items on artificial life, chaos, evolution, etc. All of them delve into this concept of emergence: this notion that totally unexpected - but organized - behavior can result from the complex interactions of very simple rules.

    The peculiar thing about emergence, though, is that all these authors and researchers are fascinated by it but nobody seems to have a rigorously measurable definition. Nobody seems to be able to say that "if and only if X and Y happen, then you have emergence."

    Has anybody here seen such a definition?

    • From what I got from this book, (not a definition, but perhaps an explanation)
      If you have many simple units preforming behaviours built on simple common rule-sets, who all interact together, and react to these interations according to their rule-sets, and a macro-behaviour occurs (something that none of the individuals would be able to do by themselves) this is Emergent behaviour.
    • Nobody seems to be able to say that "if and only if X and Y happen, then you have emergence."

      That's the thrill of science. (And most of it...) Something nebulus on the horizon like, the seven cities of gold, you know its out there but you can't quite put your finger on it, so you cast out on various paths to see what you come up with.

      Emergence is very new, as in &lt30 years old for really serious studies (I'm sure people were contemplating it before then, but not really analyzing it at all) I mean just as short as 100 years ago we still thought simple math would solve every problem.

      This particular branch is exceptionally interesting because it seems to have character in a VAST number of disciplins and possible all of them. That the character of emergence, if boiled down, could be of pragmatic use in all sciences. Very cool (not to mention everyone's guess about emergent intellect.)

      Nathan.

    • I'm currently readin Kauffman's "At home in the Universe" [amazon.com] [amazon.com]. And he's exploring under what conditions complexity emerges from seeming chaos. The book isn't exactly lite reading, you ought to have a more than passing understanding of probability & chemistry, among other subjects.
    • I just today tried to explain emergence to a crowd, and failed. What I did not tell them, which could have succeed but make matter less interesting for them, is where redecutionism fails, emergence prevails. It is about state of art, rather than an insrinct property of the system. If you can say "rules X, conditions Y makes behaviour Z" there is no emergence to speak about. The concept of emergent property has nothing to do with complex systems analysis, when whole is more than sum of its parts, the extras in the whole are said to be emergent. Ofcourse if you really know how to sum things, the whole is always sum of its parts. Water's properties were emergent from hydrogen's burning once, since water had nothing in common with either hydrogen or oxygen. Now that Schrodinger's equation can be solved for the system, water's properties are no more emergent, they are summation of hydrogen and oxygen; only that addition operation involves not simple addition but a system of differentials. The same idea applies to emergence in complex systems, but I don't have a ready example for emergence->reduction in those systems.
  • Theologians work for centuries to finely craft their ideas about God.

    And the best you can come up with is, 'He's the Mayor from Sim City.'

    I hate the fact that scientific papers never say, "God did it. We don't have to bother with this anymore, it's just too complicated."

    • I hate the fact that scientific papers never say, "God did it. We don't have to bother with this anymore, it's just too complicated."

      It's because scientists have the perseverance, and the _belief_ that human reasoning is a sufficient tool for obtaining any knowledge, even the ultimate truths. IMHO religions based on 'fixed truths' are a reflection of mental laziness. When I see people sweeping problems under the rug, just by saying 'it's God's work and we should leave it as a mystery', I get a feeling that they are (a) afraid of truth, or (b) lacking the self-confidence to go out and explore the problem. The latter can easily translated as mental laziness, a closed shell of comfort.

      In effect, if you believe that in the end it is a mystery which should be left in peace, you should not even start any scientific study, much less write a book about it. I think every question 'Why?' will eventually trace back to one single question: 'Why does the universe exist at all?' If you can say 'Because God did it', then the same answer can be applied to any other scientific problem. Which of course means that there would be no science.

      Let me emphasize again: Science is all about optimism, the belief in yourself - that we as humans have the mental power to understand nature. As such, this is in no contradiction with Christianity or similar religious ideas. But with enough confidence, who needs a god anyway? ;-)

    • While I do believe in God I can't agree with your post. In my opinion God and science can and do co-exist. Just because God did it doesn't mean people can't/shouldn't try and understand it. Because of people trying to understand science many diseases have been cured along with a number of other very positive impacts on humanity. God also did another thing, and that is giving humans a mind that can think and figure these things out, why would God have done that if he had no intention of us learning about the world. Also while I do believe in God you have to recognize that other people may not, and although I don't agree with them I do agree that it is their choice.
  • by zondance (69315) on Thursday December 13, 2001 @12:02PM (#2699384) Homepage
    I heard this author on my local NPR station [kuow.org]. You can hear the archived .rm here. [128.208.34.102]

    --Zone Dancer

  • Ilya Prigogine (Nobel Laureate) extended a lot on that matter years ago. I *think* he is one of the first writing about 'emerging properties' of complex systems. Have a look at "Order out of Chaos", or "The End of Certainties", just to name a few... Nevertheless, I think this is one of the most fascinating branch of fundamental research of the last 20 years. This also relates to an article [slashdot.org] publisher here on Dec 4...
  • For a variety of reasons, I am very interested in emergent behaviour. However, most of the literature about it has been too focused, applying only to a very specific field and without any effort to extract general principles.

    Bioligists discuss it in terms of evolution, physicists talk about it in terms of particle physics and cosmology, sociologists talk about it in terms of city formation and anthropologists in terms of cultural dynamics. But none of them try to explain emergence in and of itself, except for the mathemeticians (and good luck figuring out what the hell they are talking about if you don't have a higher degree in pure math).

    As a prior commenter stated, many people *talk* about emergence, and derive emergent principle from their disciplines, but very little effort is made to define what emergence is, how it works, and how to tell the difference between emergence and random noise. In answer to another commenter's question, emergence frequently *is* a matter of scale, random behaviour at one level displays emergent properties at the next level, then reverts to apparent randomness at higher scales, and then can display randomness *again* at the next jump in scale. Emergence is what happens when all the "random" fluctuation happen to cancel each other out in just the right ways to make something definitely non-random happen.

    That's why this book is useful, because it examines emergence as a whole concept in plain english, rather than just focusing on a particular example or type of emergence.

    --Dave Rickey

  • I thought that is was one of the best, most thought provoking books I've read in a long long time. I was able to find a copy when it was in publisher pre-release, and I found that 'Emergence Theory' was exactly what I've been looking for as the glue to make me understand the idea of genetic/natural computing and how it relates to evolution.
    You can see who is visiting your site, unless they are using an anonymizer proxy, or other system to hide your headers. The HTTP-REFERRER header gives you exactly this information.

    I thought about the fact that one could look at the referer data from a web server to make http bi-directional, but I think that for any kind of emergent behaviour to come about, one has to have many individual units all behaving in similar ways, so even if one or two people were to hook in some smart web scripting that looks at, and reports who refered to them, it wouldn't make a difference unless a large critical-mass of individuals did the same thing. So HTML over HTTP in it's present incarnation is not bi-directional for all practical purposes.

    The reviewer doesn't seem to like the idea of micro evolution without the intervetion of God,

    There is no mention of where these rules as such come from except through evolutionary survival or initial chance...
    The idea that a God figure could be there, tweaking the parameters as the model runs, or even setting the initial conditions works against his ideas.
    Could be, but this is a book based on scientific theory, so you can't really blame it for not showing a creationist side of the argument.

    More than anything, this book made me want to go out and write some cool internet client/server that would do some very simple, known, micro-behaviours but if distributed widly around the internet, and interacting in various ways, they would produce, undefined, Macro Behaviour! That is what Emergence Theory is all about, if I understand it correctly. Same thing that's happening in the brain, Neurons are doing very simple micro-behaviours, but put them into a great big hunk of brain where there are millions of simple units interacting and doing micro-behaviours (fire-nofire) extreamly intelegent and unpredictible Macro-behaviours occur.

    This book got me all excited because this was the first time this was explaned to me in a simple way using analogies that I understand, like software, games and even Slashdot! I recomend it, Steve Johnson is also the author of "Infomation Culture" which is apparently a pretty good book too, I havn't read it yet.

    yebbNOSPAM@kidojo.com

    • On the subject of reverse-linking through combing through your logs, it doesn't seem like this would be a particularly difficult problem, except that it would need a different parser for each web-server implementation. To reach critical mass, you'd need to find a subcommunity where for whatever reason most of the sites for it used the same architecture. However, for focused community sites, it would seem a natural thing to do, it surprises me no-one ever tried it.

      As for the reviewer's angst about the failure to explain God's role, I'm surprised the same thing didn't jump out at him that did at me: Emergent systems are controlled from the edges . Direct intervention rarely produces the results you are looking for, you have to monkey with the fundamental rules.

      God's all-powerful and all-knowing, but he worked himself out of a job in the first planck unit after the Big Bang. Since then, there's been nothing for him to do but sit back and watch it run.

      --Dave Rickey

  • There's another book called Complexity (1992 - M. Mitchell Waldrop (Amazon link [amazon.com])) which is a great early history and overview of the field of complexity and how it was formed (and which emergence is just another name for). Nine years old, but still great reading. Includes the formation of the Sante Fe Institute.
  • by mdecerbo (9857) on Thursday December 13, 2001 @12:26PM (#2699531)
    I got excited when I saw this review, because I thought people had finally noticed the other "Emergence", a really great sci-fi book [homestead.com] which earned author David R. Palmer two Hugo and one Nebula nominations.

    It's hands-down the best post-holocaust SF I have ever read, but it is, incredibly, out of print. If you like this sort of SF, it's worth tracking down a copy.

    Unfortunately, the author wrote one more book, "Threshold", and then disappeared entirely. I don't know whether he passed away, ditched writing, or what, but it's a shame.

  • by tcyun (80828) on Thursday December 13, 2001 @12:33PM (#2699582) Journal

    If you are looking at some additional texts in the area, Dr. John Holland has written two books. (Holland is also a MacArthur award winner, which places him in some fairly good company.)

    - Emergence : From Chaos to Order (Helix Books)

    - Hidden Order : How Adaptation Builds Complexity

    I thought /. reviewed one of the books earlier, but a quick search did not find anything. As I recall, Emergence is the earlier book of the two and is much more technical. Hidden Order is more topical and discusses concepts as opposed to technical details... but it has been a few years since I read either.

    Just some info for those who might want another angle on a similar subject.

    • by Anonymous Coward
      Emergence was the second book but still a great intro to the topic. He takes two examples - Arthur Samuel's Checkersplayer program and neural networks - and uses them as a foundation for discussing the concepts behind emergence.

      This book is well though out, fairly well organized and avoids using 'populist' examples in favor of examples that easily demonstrate the concepts without stretching the imagination.

      -Chris
  • EGB (Score:4, Insightful)

    by Untimely Ripp'd (513408) on Thursday December 13, 2001 @01:03PM (#2699739)
    Jeez, don't geeks read Godel, Escher, Bach anymore? Hofstadter did the ant thing 25 years ago.
    • Took the words right out of my mouth. Geeks! Go read GEB! Ant-colony/human-brain analogies aside, it's so much cooler than all these recently published books on desperate quests to find a Grand Unified Theory of Everything, Ever. Process rather than product. Plus, you get to break number theory.
  • So, like most popular science books currently available it will educate you, entertain you and keep you occupied while reading it or totally bore you. But it is not a book of philosophy to base life on, which thankfully, the author has not tried to provide.

    Or, with equal logic, "it is not a book of programming style and architecture to base applications on, which thankfully, the author has not tried to provide."

    What's the point in saying that it isn't something it doesn't claim to be? So you were hoping for support of your personal beliefs and didn't get it? To bad. That isn't a reason to go from saying that the book will educate the reader (in the quote above) to calling it fiction:

    I really wanted to like this book. But the level of information within it will make me put in into the light, popular fiction section of my bookshelf.

    StartRant:
    A belief system must be pretty desperate for support when a science book gets relabeled to fiction, not for an attack, but merely for failing to provided the support the reviewer had hoped for.
    :EndRant

  • You can't really know what pages are pointing to you by looking at referral headers.

    Referrals only tell you what people came to your site _from_ a said site, not if a link exists or where the link is.

    Think about:
    - use of a link redirection page by a site (usual in corporate sites, etc)
    - links that exist but aren't followed
    - people that surf with referrals turned off (lots of software allow this)

    The link system is effectively one-way only.

    Two-way links were proposed in the original Xanadu system, but no provision exists under the current www scheme. Except, of course, for google... :)
  • by Jim McCoy (3961) on Thursday December 13, 2001 @02:11PM (#2700151) Homepage
    It seems that the review of Emergence has about as much substance as the book itself, collecting random bits from a larger body of work to prove an almost unrelated point. A reviewer who finds the lack of a god figure in a book about emergent behavior unsettling? That is the whole point of complex adaptive systems you idiot! Rich and varied macro behovior arises from simple rules applied at the micro level in a massively parallel fashion.

    That is not to say that Emergence is a good book. It is an adequate book to give to a lay reader who is completely unfamiliar with the subject matter so that they can at least understand the basics of emergent behavior. On the whole the book is about at the same level as Kelley's Out of Control, cute but nothing of consequence. Anyone who is really interested in this subject should start with the following list:

    Turtles, Termites, and Traffic Jams (Michael Resnick)
    Emergence (John Holland)
    Hidden Order (John Holland)
    At Home in the Universe (Stuart Kaufman)
    A Self-Made Tapestry (John Ball)
    Swarm Intelligence (Bonobeau et al.)
    The Computational Beauty of Nature (Flake)
    Anything (and everything you can find) by Dawkins, E.O Wilson, and Hofsteader along with the Artificial Life series from the Sante Fe Institute (preceedings from the conference series of the same name)

    This is an interesting and important subject area which most Slashdot readers would be well-served to examine and explore. Unfortunately such exploration is not served well by either this review or the book being reviewed.
    • Agreed. Incidentally, I have high hopes for A New Kind of Science [wolframscience.com]. Undoubtedly, it'll contain more than its share of speculation. But, hopefully, the majority will be well-supported by verifyable experimental results [wolframscience.com].
    • Agreed,

      Emergence is an adecuate book for someone who does not know anything about the subject. For such a profile it could be big eye opener. However for those who were expecting more concrete matters, it was shalow and lacking substance.

      I can recomend "The Timeless Way of Building" by Christopher Alexander. Even when this book talks about architecture, it touches in very insightfull ways on emergence, patterns, and self orgaization.



      F.
  • I read this book when it first came out and I've been working on a review of it myself. I figured it was an ideal candidate for a review on /. given Steven Johnson's (the author) multiple referrences to /. in the book. It is interesting to note Johnson's background in the context of site many Slashdotters used everyday. Johnson was a founder of now dead community generated content sites Feed Magazine [feedmag.com] and Plastic [plastic.com], which are very similar to /. in the way they are generated an community maintained. Plastic even uses Slash [slashcode.com] as its base. I found the sections pertaining to how sites like these work to be very insightful and they'd probably be of interest to anyone who's ever wondered why /. works as well as it does.

    Additionally, our reviewer leaves out the parallels between biological emergent systems (slime molds, termites, etc.) and computer systems. Johnson gives an entirely new deconstruction of the 'pacemaker' or 'queen ant' theory in both computer and life systems. Altogether, I think the book is worth the 3 hours it takes to read.

  • "The Origins of Order" [santafe.edu] by Stuart Kauffman is an excellent book on this subject. It's pretty heavy reading, plenty of math and experimantal models, with a focus on biology and how the order we see in living systems arises out of chaos.
  • Those interested in current discussion of various models involving emergence and their specific use (or uselessness) in comprehending consciousness should check out The Emergence of Consciousness [imprint.co.uk] - which is the book version of the Sept.-Oct. 2001 issue of the Journal of Consciousness Studies [imprint.co.uk]. Robert Van Gulick's overview of different types of emergence which have been theorized is particularly valuable.
  • The principle of emergence is that interestingly
    complex behaviors results from simple underlying
    phenomena. For example Mandlebrot fractals are a
    six line program in FORTRAN. Human mental activity
    arises out a trillion elementary nerve cells, etc.

    The problem is, this is unpredictable.
    By definition science is repeatable experimentals
    and observations.
    So emergencent phenomena are not predictable until
    that happen, and therefore non-scientific.
    • There is scientific research being done in the field, primarily focused on trying to understand the basic principles of complex systems. Many make comparisons between the fields of complexity now and thermodynamics 200 years ago. In time, we will have lots of new equations and laws about this stuff.

      However, you are correct in that most of the examples you see in cited in books are just toy models. Interesting, but they don't tell us anything useful about complex systems.
  • by sunhou (238795) on Thursday December 13, 2001 @07:41PM (#2702066)
    Another book on the topic that came out probably almost 10 years ago is "Emergent Computation", edited by Stephanie Forrest. It's out of print now, but I believe it was also published as a special issue of the Physica D journal. It was a conference proceedings. (I used to work at the Center for Nonlinear Studies and Santa Fe Institute, and Forrest was also around at the time.)

    By the way, people interested in this stuff may be interested in checking out the Swarm [swarm.org] simulation system, a multi-agent simulation environment. Some of the demos that come with it are the ant/pheromone models and so on, which e.g. Resnick also explored in StarLogo.

Let's organize this thing and take all the fun out of it.

Working...