Talk:Emergence/Archive 1

This is an archive of past discussions about Emergence. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 1

Archive 2

Archive 3

Moved from Talk page of original "Emergent behaviour" article (Lexor 22:31 15 Jul 2003 (UTC)):

please* don't link every second word of this... it will become unreadable...and consider this a draft - literature needed! -- OlofE 21:57 Sep 23, 2002 (UTC)

Water example

The water example is interesting, but I don't think it really qualifies as an emergent property. The sense of emergence is that the property emerges from a large aggregation. Two hydrogen atoms and one oxygen atom can combine to form water, but the properties of water do not emerge from the combination. The combination is actually a radically different thing, due to the quantum and wave characteristics of the electrons involved.

The ant example is the perfect example. If you have two or three ants, there is not much change; they may interact, they may not. If you add another one or two, still not much change. Twenty more ants, still not much change. If you keep adding ants until there is a large colony, then the colony will have a behavior that is at a higher level of description than the behavior of the ants themselves, who will continue to exhibit many of the same behaviors as in the past. The properties of the colony's behavior will be different from the aggregate behavior of the smaller groups of ants. Hu 19:49, 2004 Nov 17 (UTC)

I agree. The fact that corrosive oxygen and combustive hydrogen make water is exceedingly cool, but it is not an emergent phenomenon. I'm removing this. Jmeppley 22:11, 18 Nov 2004 (UTC)

Disagree about the ant example. Is not the behaviour of the ants in large part defined by the fact that there is a child-bearing queen, which worker, warrior and drone classes exist merely to serve, protect and perpetuate? Isn't a colony not a colony without its queen? Perhaps an entomologist can help us out here.... HarmonicSphere 3:46, 26 Apr 2006 (UTC)

The water "example" was re-entered by another anonymous IP address, without comment or explanation here. After a little thought I removed it, per this discussion. Hu 00:49, 2004 Nov 22 (UTC)

A reference supporting water molecules as an emergent property is Sole, Richard and Goodwin, Brian (2000). Signs of Life: How complexity pervades biology. New York: Basic Books. Peer 15:40, 10 December 2006 (UTC)

External Links

The external links are getting out of hand. First, the links should be useful text, not just a url. An interview with Steven Johnson, author of Emergence not just http://www.oreillynet.com/pub/a/network/2002/02/22/johnson.html. See Wikipedia:Describe_external_links

Second, my sense of what wikipedia should look like is still forming, but I don't think it should be a clearing house for all related links. I don't think we need to link to every online book or story about emergence. I'm removing three links. http://mind-brain.com/emergence/ doesn't link to anything. http://www.kk.org/outofcontrol/ch2-b.html is already mentioned in the references. http://www.collacomp.com is a software company and offers no information on emergence. Jmeppley 22:06, 18 Nov 2004 (UTC)

Water explanation

I added the water example (the second time, anyway) based on a few independent factors.

1a) My Educational Background. I learned of water as an example of emergence in chemistry classes from high school through college (including my 10th grade Chem textbook, for which I can track down the info if necessary). Sorry for the lack of comment initially - I merely thought it was a textbook example and therefore uncontroversial.

1b) I did college work in systems theory and remember a few resources (Britannica being the most prestigious) that use water as an example. I checked with Britannica Online after seeing critiques here, and their visible excerpt includes water.

2) Change. The very notion that "the combination is actually a radically different thing, due to the quantum and wave characteristics of the electrons involved," is important. It means that the property of anti-flammability literally "emerges" from a specific electronics of water. This property is not, as we can all agree, fundamental to the constituents separately, or even a property of a mixture of the constituents.

3a) Fractals (The book Complexity by M. Mitchell Waldrop is most straightforwardly thorough on this topic). While I appreciate Hu's critique (lack of aggregation of interactions in the water example), I think Hu's definition confuses the fractal nature of complex systems with the more fundamental notion of emergence. Emergence is an example of any single iteration of a fractal (or complex) interaction. Complex systems are systems in which fractal interactions occur. Therefore, an emergent result of "a specific interaction of hydrogen and oxygen molecules" is water. An emergent result of "water molecules interacting with each other below freezing" is its lower density as a solid than as a liquid. Each step listed is emergent and complex. The aggregated result ("hydrogen and oxygen molecules specifically interacting" is "ice's lower density than liquid water") is rather a complex system consisting of multiple iterations of fractal interactions. This aggregation of iterations is complex, while each iteration is emergent (and, as a result, complex).

3b) In other words (using a simple perfect 2D fractal), emergence tells us 1 + 1 = 4, OR 1 + 1 + 1 = 16, OR 1 + 1 + 1 + 1 = 64, etc. Complexity tells us 1 + 1 = 4, AND ALSO that 1 + 1 + 1 = 16, 1 + 1 + 1 + 1 = 64, etc, as we add levels of interaction. Complex systems are, typically, fractal interactions in aggregate. The individual iterations of interactions are complex; but fundamentally, they are emergent. Why is each level emergent? Because 4 is not the sum of 1 and 1. It is the fractal result of 1 and 1 interacting. The extra sum literally emerges from the way 1 and 1 interact. If it emerges, it is thus emergent.

4) Scope of Change. With all due respect to Hu, the following comment is technically valid, but not relevant to emergence. (Admittedly, it is vaguely relevant to complexity as a field of study, in terms of self-organization... But our focus is narrower here.)

"The ant example is the perfect example. If you have two or three ants, there is not much change; they may interact, they may not. If you add another one or two, still not much change. Twenty more ants, still not much change. If you keep adding ants until there is a large colony, then the colony will have a behavior that is at a higher level of description than the behavior of the ants themselves, who will continue to exhibit many of the same behaviors as in the past."

If we remember the butterfly effect, the amount of self-organization to create a tornado/tropical storm is not dramatic UNTIL the number of interactions becomes large enough that one iteration of seemingly minor interaction drastically changes that number. Can remove the butterfly because it is apparently insignificant? No. The fact (of fractal interactions) matters more than the scope of results during any one iteration.

4a) Emergence is when the result of actions is not inherent in the actions or the actors. This is true with the butterfly effect. It is true of water. It is also true of the ant colony. However, the scope of the change only matters in complex systems when discussing rate of self-organization. The fact of the change (regardless of rates) matters in emergence. Nevertheless, the scope of change is still significant when we address H₂ and O₂ being flammable, and H₂ being anti-flammable. In one single iteration, there is a significant change in properties. So, while Hu's premise fails, water still fits its test.

Hu's Test

Now, if I were to agree with all Hu's premises, water still applies. I will borrow Hu's phrasing to illustrate how. Original quotation: "The properties of the colony's behavior will be different from the aggregate behavior of the smaller groups of ants." Rephrased: "The properties of water's behavior will be different from the aggregate behavior of the groups of oxygen molecules, hydrogen molecules, or a mixture of hydrogen and oxygen molecules." This is true.

True, perhaps - but irrelevant. By this definition, almost every chemical reaction is emergent behaviour as the properties of the resulting compounds/molecules/whatever are often not predictable from the properties of the ingredients (e.g. sodium + chlorine = sodium chloride (salt). You *might* get away with claiming that water displays emergence because the properties of one water molecule on its own may not allow you to predict or extrapolate the properties of a congregation of water molecules - but I doubt it, because many of water's macroscopic properties can indeed be deduced from the structure and behaviour of a water molecule in isolation.152.91.9.115 04:47, 3 February 2006 (UTC)

The properties of water's behavior are different from the aggregate of all the hydrogen, or all the oxygen, or all of both. Differently approached, let's ask if water fits each leg of Hu's test: Does something emerge? Yes. Is the thing that emerges different from the smaller groups of the constituents (or even, in water's case, the constituents together)? Yes. Is the emergent result significantly different in terms of properties not exhibited by the constituents (or their aggregates)? Yes. So, if Hu's premises were applicable to emergence, water would still fit. Emergence is more basic, so we can be assured that water belongs.

The Basic Point

As we focus on emergence (that single instance in which something "emerges"), we can easily ignore the aggregate of interactions. That said, this doesn't mean the ant colony example is wrong. The ant colony slowly becomes more complex at first, as interactions between ants aggregate; but the aggregation (of interactions) is simply re-iteration (fractally) of emergent interactions. There is a "tipping point" where the number of ants becomes so that self-organization occurs rapidly, but speed and scope are not relevant to that fact.

The long and short of it is that complex systems depend on fractal interaction, whether one iteration or more. Emergence is the single instance of a fractal iteration. Or, emergence is when the result of actions is not inherent in the actions or actors themselves. Or, emergence is when the result is more than the sum of its parts.

I apologize for the length of this message; I simply want to be thorough. Also, I do not want to re-add the water section without response, so I will leave the section as it is. I look forward to a hearty discussion.

Zenter 2004 Nov 22 (UTC)

Thanks for the feedback. I still haven't digested it, but I'll try to explain my position in more detail than in my earlier, slightly snarky (sorry), comment.

My background is physics/math and I recently started working in ecology. My initial introduction to emergence was in a class on complex systems that used the Stephen Johnson book as a text book. The Impression I get from him is that emergence is a property of systems with many simple components. Usually the components are all identical or extremely similar. Usually, the components are connected loosely, the total effect of many components multiplied by many weak interactions between the components results in emergent, self-organizing properties. So, I would agree with Hu. But I used words like "usually" because I'm not 100% sure of the formal definitions here and I associate self-organization and emergence.

The one comment I'll make at this point (with limited research so far) on Zenter's post is on the Britannica example. From a quick survey of online definitions of emergence (google: define emergence) I noticed that a few places did use water as an example.

http://smile.jcon.org/soft/product/ref/entry/e/Emergence.html

and I also found the following:

http://www.masterliness.com/a/Phase.matter.htm

These examples speak to liquidity (and phases in general) as an emergent property of water that is not apparent when only a few molecules are present. This is a very different concept than the example that started this discussion. It is not clear to me which of these phenomena the Britannica article is referring to.

For me, a key component of emergence (in the complex systems sense) is that the individual pieces aren't changing, their actions just sum differently when there are more than a critical number. If you have 20 water molecules, each molecule is indistinguishable from any water molecule in a teaspoon of water, yet the fact that there are so many molecules in the latter case, changes the way the system behaves. On the atomic scale there is a difference between an oxygen atom in water and an oxygen atom in O₂. Jmeppley 19:52, 22 Nov 2004 (UTC)

I apologize for the length of this message; I simply want to be thorough. Also, I do not want to re-add the water secton without response, so I will leave the section as it is. I look forward to a hearty discussion.

Zenter 2004 Nov 22 (UTC)

Hu writes: Thank you Zenter for a detailed response. My interest in emergence comes from my interest in intelligence, artificial intelligence, and consciousness; so I will move to discussing that as an example. In my understanding of this, emergence is the jumping up to another level of description beyond what was sufficient to describe the pre-emergent system.

In the case of the formation of water, the level of description remains the same, i.e. physical and chemical. The properties of oxygen, hydrogen, and water can all be measured with variations of the same instruments. The fact that two are gases and one a liquid at room temperature is just a matter of degree (literally!). The flammability of the hydrogen (when in presence of oxygen) and the non-flammability of water are just stations along a dimension of qualification that pre-exists. Some elements and compounds have flammability, others do not. The non-flammability does not so much emerge as it is acquired. Another strike against the formation-of-water model is that the units are changed. The nuclei may be the same, but the electron clouds are radically different.

If we look at the action of uni-cellular organisms like amoebae, we see the basics of animal life: seeking and consumption of food, avoidance of harmful environments, and reproduction. As we increase the complexity of the animals up through invertebrates and vertebrates, the aggregations of cells exhibit more complex forms of the behaviors, but fundamentally the animals still seek and consume food, avoid harm and reproduce. However when we arrive at human beings, new levels of description are necessary. Descriptions like "government", "religion", "wiki", "superstar", and "music" are just a few of a myriad of new concepts that can be applied. These are dimensions of qualification that do not exist in simpler animals. We can say a chimpanzee does not have religion only because non-human animals do not have religion. Religion was never an issue until the human species arrived. There is nothing inherent in human brain cells compared to chimpanzee brain cells, the units are essentially the same, but yet in the transition, something dramatically different is emergent. Hu 22:47, 2004 Nov 22 (UTC)

Zenter writes: Hi... In composing the responses, I spent some time sitting on them, re-writing them, thinking some more... Sorry, again, for the length.

I want to take a couple steps back. We are not defining our terms/perspectives/assumptions before plowing forth. (Ex: I don't understand the precise meaning of "levels of description," nor its relevance)

My Perspective

The reason why I studied systems theory in college was because I majored in religion, specifically Buddhism. In all Buddhist schools, there is a rejection of souls. This is because of various reasons, most important being that an unchanging eternal "self" is affected and makes decisions by the changing world, and thus cannot be unchanging, and thus it cannot be an eternal, unchanging self.

Mahayana Buddhists took this a couple steps further, rejecting "selfhood," or essence, of anything in the universe (including themselves). Basically, they were rejecting a philosophy similar to Platonism. Their argument is logical reduction: where is the "tableness" within a table? Is it alone in the flat board? Or the legs? No. The "tableness" comes from multiple sources: the board, the legs, and me working together. The "tableness" arises co-dependently. If I sat on a table, I negate its tableness. It now has chairness. However, if I chopped this table in half, so it could not stand, it would have neither tableness nor chairness. The form thus is important, as is my use of it.

The same question is asked of a river (what is a river? where is its riverness?) and fire (what is fire? where is its fireness?). The result is always the same: an uncertain place somewhere in between all the actors working in concert. The lone table, in its own universe, is no more a table than I am a person in my own universe. Mahayana Buddhists typically call that unstable/uncertain interdependent creation "dependent arising" or "dependent co-arising."

Having learned a great deal of science, I immediately saw a connection between this philosophy and the notion of emergence. Heck, I thought, the word "arise" and "emerge" are synonyms! Emergence, therefore, is when something arises co-dependently. Predictability and statistical probability were less important than the simple fact of interdependent arising. It is predictable that when I hit a pool ball, it moves. That does not mean, to me, that is not emergent. It simply is a really boring instance of emergence that depends on me, the pool ball and the cue. Bu, just because something is boring or predictable does not mean it is not emergent.

What is Not Emergent?

So what's not emergent? Independent events. Let's say I support a team that makes it to the championship. Let's also say that at around the same time, there is an important election. I want my team to win, but I REALLY want my candidate to win. So I make a deal with the fates: I will sacrifice my team for my political candidate. My team ends up winning. Does that mean my candidate will automatically lose? No. When my candidate loses, can say that this was an interdependent interaction between myself, my team, and my candidate? No. That my team won and my candidate lost are independent events that I connect. So, in my mind, there is a connection, but the two events have no bearing on each other. That is because there is no interaction.

Basically, to me, emergence is what happens during an interaction. There is some sort of exchange. This conversation we are having is emergent. It depends on you, me, common language, common interests, and wikispace, among many other factors. All these factors meet in an uncertain creation of a conversation.

It reminds me Schrodinger's Cat. There is a level of uncertainty built-in to all sciences, and however minute that level of uncertainty is, it exists. That an interaction results in something that was not "there" before is enough to warrent the label "emergence." The reason is, emergence is (to me) the most basic building block of all sciences, and I think the definition should be extremely inclusive.

What is Emergent?

What is emergent? Almost everything interactive to the point of mundanity or uselessness. Why? Like the butterfly effect, a mundane interaction can have non-mundane effects. Let's take a snippet of a gene of a made-up harmless bacterium.

ATTGGCAGGTACC

Now let's make a point mutation.

ATTGGCAGGTGCC

This mutation was a result of random interaction with basic light radiation. The result of such mutations is usually harmless, and therefore, usually not worth our attention. But here, however, we get a deadly bacterium that attacks the lungs and is also resistant to most forms of antibacteria. This new bacterium is the emergent result of a random-chance interaction. But so are ALL the harmless ones. We cannot base our label of"emergent" on whether the event is significant (to us) or not. This was my previous argument regarding "scope."

An added issue is entropy. My understanding is that emergence moves against entropy. We take hydrogen and oxygen and a lightning bolt, and these three entropically stable things organize into something else with different chemical and physical properties, against entropy. If it moves against the flow of entropy, it is complex. As for self-organization, I suppose it stands to reason that as long as there isn't widespread self-organization, it would be difficult to argue in favor of a single point mutation, single molecular reaction, being emergent. I guess my perspective is a little less rigorous - I am not forced to ask, "Is this significant enough?" But the upside of my perspective is that there is less judgement and, more simply, a switch - "Does something change versus entropy?" I am trying to keep "me" out of the definition as much as possible, because "I" bring a perspective and biases, while a switch does not.

Which leads me to...

Response to Hu, Jmeppley

I want to thank Jmeppley and Hu for taking the time to explain their points of view. I think I agree (or at least understand) where you're coming from, Jmeppley. Most of my disagreement is with Hu, it seems. Any disagreement I have with Jmeppley is embedded in my larger rebuttal to Hu.

Hu, I agree that human concepts/conceptualizing is emergent. I also think the cooperation of various structures (from organelles to organs) within a living thing is something that leads to an emergent "single being." "I" cannot be without my heart, nor my heart without "me." The heart is not "me," but there is no "me" without the heart. This interdependence is a great example of emergence. The heart works in a specific manner in relation to other things, seemingly independent. The result, however, is a seemless, single, "me." This is consistent with your "city" example in the definition. I also think this is consistent with my "water" example. (O₂ works in relation to 2H₂ to result in a two single, seemless, H₂O molecules.)

Anthropocentrism

As I said above, I don't understand the relevance of "levels of description." I also do not know why observation based on the same schema (ie physical and chemical properties) is automatically dispensed as "not emergent." I don't see the form of observation as crucial to emergence. The fact that something occurs interdependently to create a result that is unique is, to me, the essence of emergence. That it is unique in the same way other things are also unique does not make it any less emergent. I feel like this is a step to "value-judging" events rather than keeping human value out of the discussion.

For example, Hu says: "As we increase the complexity of the animals up through invertebrates and vertebrates, the aggregations of cells exhibit more complex forms of the behaviors, but fundamentally the animals still seek and consume food, avoid harm and reproduce. However when we arrive at human beings, new levels of description are necessary." Are humans really that special? This is dangerous anthropocentrism, from my perspective. I would lump the fact that we have religion and government in the "consume food, avoid harm, reproduce" category, as long as we are lumping things like organ specialization, immune responses, and the fact of sexuality into that category. The differences in "level of description" is a construct of the mind, not of reality. All human behavior can be reduced to the same things that all living behavior can be reduced to. We simply have greater insight on human behavior. But, if religion and government are special, why are they more so than a birds flocking, or whales using sonar? Where would the fact that juvenile bottle-nose dolphins masturbate fit in? Or that many invertebrates have a two-stage life cycle?

There's an implicit argument within your larger argument, Hu. It is that humans are, in essence, the "greatest" creatures, the rulers of the world, sort of the "end of evolution." It seems that other evolutionary changes are "not emergent," that humans are a leap above other creatures. In some old religions of the near east, the universe was arranged this way, with less similar creatures at the bottom, and we can work our way up to animals closest to us. Then there is this huge gap between the most "complex animal" and humans, often described as "a gap as big as the one between gods and humans." We humans are the intermediaries between the created world and the cosmic divine. Obviously, evolution and the development of all the flora and fauna on earth was more complicated. So, while I agree that religion and government are emergent, it is not because they are special social structures overlayed on top of survival. Indeed, they are a result of and purveyor of survival, like other things that have arisen through interaction of creatures (such as flocking, sexuality, and immune response - results and purveyors all, of survival).

This sort-of "judgement" in your (Hu's) line of reasoning scares me. Emergence is a property in the same way evolution is a property - there isn't a distinct line where we can suddenly say "now it has evolved." There are small steps, each significant to the process, each building on the last (see my previous fractal argument). I am not the judge of what is "worth it," and neither is Hu or anyone else. Until "level of description" is defined in a way that meets scientific rigor (in all sciences), I am extremely wary of using it scientifically. I would also warn against the dangers of dismissing something as insignificant or irrelevant - it is that sort of thinking that allowed AIDS to become the epidemic it is.

Additional Thoughts

I have done some research to try to shed additional light on the issue. It has rather shed additional confusion. The debate we are having can be seen in the JS Mill-methane example in link 3 below. My sense of understanding is more like link 5 below.

1) http://plato.stanford.edu/entries/properties-emergent/#2

2) http://cscs.umich.edu/~crshalizi/notebooks/emergent-properties.html

3) http://www.artsci.wustl.edu/~philos/MindDict/emergence.html

4) http://www.stewdean.com/alife/emergence.html

5) http://atheism.about.com/library/glossary/general/bldef_emergent.htm

This all leads me to believe that our discussion is extremely relevant, because it actually appears to be at the forefront of complexity science. I suppose this would be a good time to call in some folks at the Santa Fe Institute.

Again, apologies for the length. This took me two months of reflection, but I think we're actually getting somewhere. I can't wait to see your responses! Zenter 17:36, 21 Jan 2005 (UTC)

My contention with Zenter at this point (you've changed my mind to some degree) is that I still don't think your water example is a classic example of emergence. I won't tell you that the properties of water are not emergent. They do emerge from the combination of Oxygen and Hydrogen.

My impression of the article is that it is primarily discussing the scientific field of emergence. Part of our problem is that the field came into to being only recently and it's boundaries are not yet determined. The initial study of emergence arose (emerged? :-P) from the realization that there were a class of natural phenomena that had defied explanation because anthropocentricly inclined researchers were looking for the control centers. In many of these cases someone had an aha moment where they realized that there didn't have to be a control center. This idea then spread rapidly and was applied to many other natural phenomena and soon people saw parallels in many other systems both natural and artifical. The question we seem to be debating is how far to extend the comparisons.

My opinion is that Zenter's water example fits the dictionary definition of emergent, but does not completely fall into the realm envisioned when the term emergence was coined. Of course, I'm pretty sure that there is not a single field of study that has not had it's boundaries shift over time. I still feel that an important component of the essence of the original idea of emergence is that the new properties are apparent only when numbers are sufficiently large. (e.g. a few fireflies in the same tree will blink out of phase with each other, but get a few hundred in the tree and the whole tree will blink in unison.)

There is definitely some synergy (wow, did I just use that word) between the study of emergence and the philosophical ideas of <<fill_in_the_blank>>ness (tableness, chairness, etc). It might not be a bad idea to add a section on the philosophical aspects of the study of emergence. Does anyone know if any philosophies used the word emergent extensively before emergence became vogue? If so, there should definitely be a section (or possibly separate page) on the philosophy of it.

I agree with Zenter that Hu's arguments are unnecessarily anthropocentric, but I believe that his heavy use of human examples comes as much from his neuroscience background as much as an innate human tendency to see ourselves as the classic example of just about anything. That the brain distinction could have just as easily have been made between rats and earthworms as humans and other mammals.

Anyway, maybe we should wind down the arguments and begin a discussion about whether or how to reinstate the article. While I would not include the example in this article if I had complete editorial control, I could not argue against including it in a discussion of emergent properties, as long as there was some clarification or concession about the non-complex-system-ness of it.

I for got to sign off anf I have one more thought. I agree with Zenter that what he is calling "emergence" is at the heart of all science, but I don't agree that that idea should be called "emergence". I think that is the root of our disagreement. Anyway, I'm really enjoying the discussion. -john

Jmeppley 18:10, 23 Jan 2005 (UTC)

Zenter writes: I agree! This is an interesting discussion. I think the philosophical "field"/perspective that basically agrees with my "dependent arising" perspective is called nominalism. And they don't use the word "emergent," sorry Jmeppley. At the very least, they concur on many points. For instance, they critique the notion of __ness, because things are defined, as they say, "in relation to." Green has no meaning except in relation to other colors.

Nominalists are not just accepted by the philosophical community, though. People who believe that there is an absolute moral (and therefore physical*) structure to the universe call this point of view, pejoratively, "(moral) relativism." They have been known to associate nominalism/relativism with nihilism, because of the lack of absolutes. In fact, nominalism is as far from nihilism as it is from materialism (or realism). This is sort of cutting-edge philosophy stuff, as it relates to post-modernism and post-post-modernism. What's fascinating is that these arguments taking place now in western philosophy are mirroring, very closely, what Buddhists argued about over 2000 years ago, which is why I find all that stuff I studied relevant to our discussion.

* A common phenomenon in philosophy is that the physical and moral worlds are dependent on one another - because if the physical world is created (by a creator), then it stands to reason that the metaphysical-moral world is created as well. If not, then morals are up in the air.

Anyway, this leads me to a couple questions: Is emergence a field? If so, exactly what does it study? As far as I knew, it was a methodological perspective, a way of understanding the systems of the universe (much like Darwinism in biology). My understanding of emergence is less strict because the idea has implications in many fields, regardless of where it started. In Schrodinger's "What is Life?", the book that basically started the fields of molecular biology, biochemistry, and genetics, he tried to define that strange anti-entropic principle of life. I think he may have used the word "emerges," but without all the baggage we're bringing! Maybe we need another name for things which go "uphill" against entropy, as you said, Jmeppley, but we don't have it yet, and I think emergence is presently the best fit. Basically, I think the dictionary definition is the best one.

Maybe a larger-scale re-organization of the page is in order. There's the simple definition at the top, including water. Then, "Emergence is most often used when studying complex systems, and it describes larger systemic anmolies which emerge against entropic principles." Examples like the early computer science "boids" experiment, ants, etc, can be cited. Then we can refer to philosophical congruencies, such as nominalism and Buddhism. At my university, my professor of Buddhism, a sociology professor (with extensive philosophy background), and a physics professor discuss the convergence of complexity and mahayana and wittgenstein all the time. If it's happening in academia, maybe we can bring that inter-disciplinary view to the wiki.

Anyway, until we decide if we re-organize, I can live with caveats. It goes against my instinct, but I'd let it go until we are able to sort out whether emergence is itself "field-dependent," and whether the page should be re-organized.

Zenter 01:35, 25 Jan 2005 (UTC)

I think we're getting really close here. I have a do have a couple of reservations. I still don't see the water example as one that gets at the heart of what "emergence" means to me. I think, judging by the language used in the intro paragraph and throughout the article, that I'm not in the minority here.

But on looking back at your original addition, it's really not so objectionable to me anymore. I first saw the second addition which added the word "classic". Your example is a good one of an emrgent property in that it is simple and easy to understand. It also underscores the point that emergent properties can be seen as ubiquititous.

Your point about emergence not being a field of study is probably true. I guess I should say that it is a key idea in complexity studies and complex systems analysis. In these fields, a size or density threshold is generally implied when talking about emergence. I can see a possible counter argument already that complex systems studies are themselves inherently concered with density thresholds. However the physics section of this article focusses on properties that only emerge for really large numbers of particles.

I agree that organization alone is enough to make everyone happy. I think your water example would be good in a section that talks about emergent properties in a broader sense, but I would object to calling it a classic example of emergence since a sizeable portion of the people who use the term regularly would not see it as such.

This is certainly a hard subject to discuss. We've been occasionally using the idea that emergent phenomena often seem to defy entropy. I was going to point out that your example doesn't seem to me to do this when I realized that entropy itself is an emergent property. Also, entropy is not a useful idea at the scale you're talkign about, so my point would have been irrelevant.

Comment on Water Explanation

The formation of water is not "against" entropy -- in fact, nothing is. Apparent anti-entropic processes seem to exist merely because the total system which is gaining entropy includes more than the "open" system which appears to be violating it. For example, the Earth would seem to have been going against entropy by the process of more complex animals and plants having evolved from simpler ones over time, if we forget that the total system includes input from the Sun, which is dispersing energy at a far greater rate than the Earth is gaining it; the overall entropy of the system is increasing.

Therefore, the whole notion of gain or loss of entropy as a defining characteristic should be junked, IMHO. Systems that approach limits in a definable way, as naturally-occurring fractals do, are often the result of oscillating processes that are winding down.

HarmonicSphere 3:42, 26 Apr 2006 (UTC)

Gaia paragraph removed

I yanked most of the recently added paragraph:

"Emergent properties arise when a complex system reaches a combined threshold of diversity, organisation, and connectivity. The property itself is totally unpredictable, utterly unprecedented, and represents a new level of the sytems evolution. An example of emergent properties on a universal scale are energy, from which matter emerged, from which life emerged, from which conciousness emerged. This is the stage that the earth has reached through us and other concious beings, such as the whales and dolphins in particular. That we have self-reflective conciousness,ie we are concious that we are concious (ditto dolphins etc), could be interpreted to be a transitional stage between the emergent property of conciousness and the next. Providing that we humans maintain our increasing diversity, organisation, and connectivity, and do not significantly degrade the diversity etc, of the rest of life on earth, then the next emergent property in the energy-matter-life-conciousness evolution will come forth. That is likely to occur on a planetry scale, and gaia will become!"

It starts out rather well (imho), but then gets awfully emotional and includes some questionable facts. I put the first two sentences in the emergent properties section and removed the rest. I'm not sure it belongs here, and if it does, it needs to be reworked. Does anyone think it should come back? Jmeppley 04:27, 22 Dec 2004 (UTC)

Zenter writes: Jmeppley, I agree that the whole Gaia thing is too much. Emergence is not so, shall we say, driven, to get some next level of understanding. It is, pun intended, more complex. I think this piece relates back to the idea of emergence having some divine purpose, which I don't trust as a systemic descriptor of the universe. Also, I'm troubled with the inclusion of "represents a new level of systems evolution." This doesn't tell me anything because I don't know what "level" or "systems evolution" mean - I am as concerned about it here as I am above, in the water discussion. It conveys a sort of "drivenness" that I would try to avoid. Zenter 17:40, 21 Jan 2005 (UTC)

Systems Thinking paragraph removed

Systems_thinking

"The most fundamental system idea is that the entity as a whole has so-called 'emergent properties', properties which are properties of a whole and are meaningful only at the level of the whole. (The vehicular properties of a bicycle are emergent properties of a particular whole entity structured in a certain way: the braking system would be a sub whole within it, also possessing it's emergent properties.) The four most fundamental system ideas are thus emergence, hierarchy, communication and control: they are the ideas needed to describe the core system metaphor"

Checkland PB University of Lancaster, UK. 1981 Op Cit. International Journal of Information Management (Vol 8, 1988, 239-248) Checkland PB Scoles J (1990) Soft Systems Methodology in Action (Wiley) ISBN 0-471-92768-6, pp309 Peter_Checkland

I decided to remove this paragraph, it looked tacked on to the end of the section on emergent systems. Especially as 'Emergence in physics' follows on well from 'The study of emergent behaviours is not generally considered a homogeneous field, but divided across its application or problem domains.' There is a link between systems thinking and emergence, but there is one included in 'see also'. This quote is probably more relevant to systems thinking. Richard Bunch

Temperature is not an emergent property!

OK, I'm sticking my neck out here! Temperature is not an emergent property. It is well defined even for a system of 1 particle in a background potential. If that particle behaves ergodically (i.e. it eventually comes arbitrarily close enough to sampling all of phase space) then it's possible to define a trajectory average of the temperature and any other thermodynamic property. This is also assuming that the particle is in equilibrium with its surroundings (which can be achieved through equilibrium with the surrounding photon bath).Christianjb 23:02, 25 Apr 2005 (UTC)

I see your point, but on the other hand you are replacing an ensemble in space (many particles at a single instant of time) with an ensemble in time (a single particle, but observed over a long period of time). There is still no notion of a temperature assigned to each particle at any given instant of time, it only emerges after considering ensembles of such particles (distributed either in space or in time). As such I would still classify temperature as emergent. Terry 02:46, 26 Apr 2005 (UTC)

Well, the average roll on a die is 3, which can be obtained either by rolling 1000 dice at once and averaging or by rolling one die 1000 times and averaging over all throws. But the average value is not really an emergent property of the dice.. or at the very least I don't think it's a very good example of an emergent property. Let's see what others have to say- but I'm not convinced. (Doesn't mean that I'm not wrong of course!)Christianjb 02:55, 26 Apr 2005 (UTC)

The average is 3.5, not 3, but never mind that... OK, I can see why the average value of a die is not an emergent property - it can be computed immediately from the state of a single die at a single instant in time. Going back to temperature, it would thus seem that temperature is a property of any system with a sufficiently "ergodic" phase space. So a particle + a potential well would have an inherent temperature, though you wouldn't be able to detect it just by looking at the particle or the potential separately, so it is still emergent in a rather weak sense. On the other hand, the laws of thermodynamics really don't apply very well to the situation of one or two particles in a potential well... even if quantities such as temperature can be defined in these situations, they don't have good properties. So perhaps it is only the laws of thermodynamics which are emergent, rather than the thermodynamical concepts themselves. I guess I should look up how "emergence" is defined in the physics literature (my guess is that the definitions are pretty vague)... Terry 18:06, 26 Apr 2005 (UTC)

Haha I have egg on my face now! 3.5 ok. Maybe I was using 5 sided dice. I still disagree however with using temperature as an emergent property. The laws of thermodynamics apply just as well for 2 particles as for 2 gazillion particles- however care must be used. It's still true that energy is conserved (1st law), entropy never spontaneously decreases, even for 2 particles (2nd law) and it gets asymptotically difficult to reach the state of zero entropy (3rd law). Now- some things like infinitely sharp first order phase changes don't exist for finite numbers of particles- but the system still approaches the infinite limit as the number of particles is increased. It doesn't suddenly become macrosopic. Christianjb 19:28, 26 Apr 2005 (UTC)

OK, I've changed the bit on temperature and thermodynamics. I hope this doesn't upset anyone because I really tried to accommodate previous points of view. I think most physicists would disagree that thermodynamics doesn't apply to small numbers of particles! If you disagree, then give me a reference. For my part I suggest looking at Professor David Wales' recent book on clusters in which he extensively discusses the thermodynamics and statistical mechanics of small systems.Christianjb 02:49, 27 Apr 2005 (UTC)

Looks pretty good to me now. Terry 20:32, 27 Apr 2005 (UTC)

MAS and QM

This may be thought of as puzzling, because quantum mechanics is generally thought of as more complicated than classical mechanics whereas lower level rules are generally less complicated (or at least less complex) than the emergent properties.

Just thought I'd defy my feeling of utter ignorence in the matter and add a short comment: It is NOT a general property of emergence to have simple agents and a complex system, though it IS for some of the main sub-groups of emergence. Exempli gratia, the study of Cellular Automata. But the opposite counts for other systems with emerging properties, as for example Multi-Agent Systems (MAS), in which complex agents build up a simple system. Or to quote the book Emergence of Complexity by Jochen Fromm, "CA are a complex system with simple agents, MAS are simple systems with complex agents.". Hence, by defining the emergence of classical mechanics from quantum mechanics as a MAS, the "contradiction" and puzzlement is obliterated.

Put that in the article! But make it simpler than the above explanation, I don't understand what a MAS is and I still feel puzzled! Christianjb 20:14, 30 Apr 2005 (UTC)

Not happy with graphics and other things on this page

The figures look nice- but they don't mean much to the non-expert and I'm not sure that they're uncontroversial representations even amongst experts. Please give references and detailed explanations of the meaning in the figures. If it's original research- then it's probably not appropriate for a Wikipedia page. If it's from an article- then please reference the original. I don't want to keep whining- but I'm pretty much unhappy with a lot of the emergence page. We need some math/physics/biology experts to help improve this page. A lot of the article is still quite flakey. I should mention that I also think there is some good stuff on the page and I appreciate the work that has gone into this article- but let's make it better! Christianjb 00:14, 18 Jun 2005 (UTC)

I really don't want to upset anyone by removing the nice figure just now. I did it for the reasons given above. If this is original research then I think at the very least that should be indicated in the fig. caption, e.g. One possible graphical view of emergent behavior. If it's not original research, then please please give references. Once this is clarified I'm only too happy for the figure to be reinstated. Christianjb 06:27, 1 October 2005 (UTC)

I am not happy with your actions, and I do not agree with your attitude. You said I am not happy with the graphics, remove them all. Instead of discussing it with others, you simply removed them, and added a termite mound picture instead. Maybe somebody else is not happy with the termite mound picture? Articles should be written without bias, representing all majority and significant minority views fairly. This is the neutral point of view policy, or WP:NPOV. You cannot say I am not happy with this, and want it like that. All point of views include the theoretical and the practical view, too. Therefore I reinserted the figure again. --JFromm 14:06, 2 December 2005 (UTC)

Hi. Firstly, I wrote the above SIX MONTHS AGO. I received no response. I waited and waited. I then went ahead and removed the graphics stating clearly my reasons on both the talk page and in the edit summary. You can disagree with my choice, but please be fair! I did things in an entirely professional manner.

I have to say I'm upset by your comments. You've persuaded me to not bother working on this page anymore. I would however like to hear others comments on this.Christianjb 14:38, 2 December 2005 (UTC)

I guess I should thank you for sourcing the picture from a copyrighted work as I requested SIX MONTHS AGO.Christianjb 14:52, 2 December 2005 (UTC)

I was upset, too. Let's call it quits. To say "I am not happy with this part" is not a valid reason to remove it completely IMHO. One should instead replace it with a better part of better quality or with a more detailed and balanced view. Nevertheless it is nice to find someone interested in the topic. I run a discussion group about complex systems (which you can find at my homepage), if you like we can discuss the principle of "emergence" over there. You seem to be very interested in this topic. --JFromm 15:21, 2 December 2005 (UTC)

Truce? OK. It's easy to get worked up about Wikipedia pages. No hard feelings. Christianjb 15:30, 2 December 2005 (UTC)

linux wikipedia and emergence

I just hacked up a paragraph citing wikipedia as an example of emergence. This isn't necessarily true. It's certainly not a great example. I think it is a great resource and a great project, I just think its not a clear enough example to make an encyclopedia article. I left the WWW as an example because it has some interesting properties (power law distribution of links and small-world-ness) that were not planned for and just happened.

Wikipedia (and most open source software projects) are amazing examples decentralized collaboration producing great results, but in most of these cases, a central group or individual exerts control over overall structure. And while, in the case of wikipedia, the content is usually not checked by the official editors, there is a set of guidelines and conventions which strive to make wikipedia great. Even if the details are left to collaboration, if the guidelines are intended to create good content, then you can't really call it emergence. Even if the guidelines have evolved over time.

Jmeppley 07:31, 23 October 2005 (UTC)

Maybe Wikipedia itself is not an example of emergence, as you say correctly. More the special "Wikipedia Community Feeling", which is quite similar to the enthusiasm for Open Source software and Linux. At least the success of Wikipedia seems to be closely related to the phenomena of emergence. Just like you say, every Wikipedian thinks "it is a great project" and "I have contributed to it" (which in turn results for some in the conclusion "I am great, too"). It is a classic form of feedback illusion, which can be found in national elections and in religious rituals, too. Although everyone has contributed only a small amount - nearly nothing to be honest -, everybody has the feeling of a giant effect through the shared feedback illusion. The possibility of instant change is important for this feedback illusion. --JFromm 15:03, 2 December 2005 (UTC)

Actually, in the way you describe it, Wikipedia sounds like an emergent phenomenon. Wikipedia does evolve, you just don't see it. Evolution, in many ways, requires time. The guidelines do evolve. If our society changes, for better or worse, the guidelines will change too. Jmeppley, your thoughts are a bit contradictory. There are officials who exert control over A PART of the structure of Wikipedia but not the whole content. For Wikipedia NOT to be emergent it would require that the leaders of Wikipedia would have complete knowledge of everything that's being written. Anyway, IMHO Wikipedia being an example of emergence is indeed nice but at the same time not helpful for the article. Since there is so much depth in this concept, simpler examples are needed, examples that everyone can see and witness, or relate to.Phoebus 13:17, 7 December 2005 (UTC)

Self-organization & emergence

I wonder is the articles Self-organization and emergence should be related more tightly? Also, should the categories Category:Self-organization and Category:Emergence (just created it) be directly related somehow? (the difference is not very clear to me) Karol 09:41, 13 November 2005 (UTC)

There is a connection in the sense that both terms or concepts are often used together because the subject system to which they refer incorporates both phenomena. However, in my opinion the concepts each denote a very different meaning. To learn more about my vision you can read a paper which is mentioned on both the Emergence and Self-Organisation page entitled Emergence versus Self-Organisation: Different concepts but promising when combined, by Tom De Wolf and Tom Holvoet --Tomdw 15:50, 14 November 2005 (UTC)

Tom, can you give me an example or two of self-organized systems in which the organization is not an emergent phenomenon? I can't think of any, but that may just be alack of imagination on my part. Jmeppley 22:58, 14 November 2005 (UTC)

In the context of computer science there are systems that can be engineered to be self-organising but not emergent. For example, when your program consists out of multiple autonomous entities that have to form a 2D shape on a 2D plane and they all have the plan or lay-out of that shape available to organise themselves. Then this is self-organising because no external control is needed to get more structure in the system. However, it is not emergent because the macroscopic pattern, i.e. the shape, is not novel with respect to the individual parts. Emergence requires that the parts have no knowledge of the whole. Of course, it may be the case that in other sciences (e.g. physics, biology, ...) there are no systems that organise themselves without being emergent. But I am not an expert on that. The big difference is that in computer science the system can be constructed completely from scratch to make such systems possible. --Tomdw 14:14, 18 November 2005 (UTC)

Question regarding emergent structure in nature explanation

Can someone explain the following statement to me?

"There is also a view that the beginning and development of evolution itself can be regarded as an emergent property of the laws of physics in our universe, or contrary to this the opposite view that the laws of phyiscs have like their constituents emerged during the course of time (in which case "evolution" and "emergence" would be the most fundamental principles in the universe)."

I think I sort of understand what the first part is saying, but then the second part makes no sense at all to me. Is this maybe some sort of reference to the anthropic principle? In any event, I think this paragraph really needs rewording. Grokmoo 04:00, 8 March 2006 (UTC)

See, this is why I really dislike people throwing around the word emergent. The first part of the sentence means, "Some people think that the presence of life in the universe is a direct, though unforseen, consequence of the physical laws of the universe." The second part does not have a clear meaning. It isn't clear what or who the constituents of the laws of physics would be (math? fundamental particles? physicists? humans?). Evolution and emergence are in scare-quotes anyway, so I think even he didn't quite know what he was saying, but it seems like what he meant to say was: "Some even think that the laws of the universe itself are an emergent property of some more basic pattern of reality." He's still incorrect in calling that opposed or contrary to the previous statement if that is what he means. -- neoshroom

Rule

Could someone please take a look at the first statement in this article and disambiguate the Rule link to the correct page. The statement is, "Emergence is the process of complex pattern formation from simpler rules." Thanks! --Hetar 09:17, 14 March 2006 (UTC)

JA: Done. Rewrite rule is the closest that we have to the sense intended, pending discussion at Wikipedia talk:WikiProject Theoretical Linguistics. Jon Awbrey 20:44, 14 March 2006 (UTC)

Video Games

Added a link to the Rampancy article, an aspect of the Marathon games (and likely Halo). Rampancy essentially comes about because of emergent behavior, so it's an interesting take on what may happen in a system (Read, AI) because of emergence. - Raveled

Color is not a physical property, strictly speaking; it's an observational phenomenon. See the color section, as linked to in the article. 71.252.18.252 03:13, 2 July 2006 (UTC)

Where's the Dialectical-Materialism here?

Why is there not one mention of dialectical-materalism in this article? That's what emergence is. 'Prior art' and all that (though this is somewhat of a rhetorical question whose answer I'm sure a number of us are quite aware of...)

Pazouzou 19:24, 3 July 2006 (UTC)

Wikipedia - an emergent system?

YES, I think Wikipedia is a marvelous example of an emergent system -- in this case, specifically of an emergent phenomena that is human caused, but not of any human design. And such a rapidly growing emergent phenomena, in my view, should be documented in a high-quality and thorough encyclopedic article on Emergence. Having said that, as the 'Emergence in Wikipedia' section evolves, I would think a brief paragraph in the Emergence article, with a link to a longer-length article on 'Wikipedia as an Emergent Phenomena' would likely be in order. N2e 02:48, 15 August 2006 (UTC)

Galaxies; an emergent system?

I see in the article under "Emergent structures in nature" the following sentence:

'Emergent structures can be found in many natural phenomena, from the physical to the biological domain. The spatial structure and shape of galaxies is an emergent property, which characterizes the large-scale distribution of energy and matter in the universe'

The first part of this sentence I feel is debatable at best. Firstly, the exact causes for the observed morphology of galaxies is not that well understood. Also, the characteristics of galaxies that we do understand by and large follow from basic physical laws and the initial conditions in relatively straightforward ways.

The second half of the sentence is rubbish. The internal characteristics of individual galaxies are not really related to the large scale properties of the universe, if I am even reading that right.

Anyway, I would be in favor of simply removing this section, but I have a feeling many will find this disagreeable. Grokmoo 19:30, 11 September 2006 (UTC)

Grokmoo, I'm not clear which of the two sentences in the quoted section you are referring to in your comment "The first part of this sentence..." Perhaps you could suggest a clarified rewrite of the paragraph here and see if anyone offers input on it. N2e 17:49, 24 September 2006 (UTC)

Hi, sorry for the bit of delay. Apparently the section I was quoting has since been changed. In any event, I made what I consider to be a fairly minor edit, basically removing galaxies as an example under the 'emergent structures in nature'. To reiterate my main point, I don't think that the shapes of galaxies are understood well enough to classify them as "emergent structures". Unless there is some debate, I'll leave it at that. Grokmoo 18:05, 27 September 2006 (UTC)

Potential Article Reorganization

With the categories of emergence examples proliferating, I think it may be time to do some reorganization of the article. Since many are editing this page, thought it would be good to discuss it first on the talk page. Current (2006-10-12) article rganization is: 1 Emergent properties 2 Emergent concept 3 Emergence in games 4 Emergent structures in nature 5 Emergence in culture and engineering 5.1 Cybernetic Emergence 6 Emergence in physics 7 Emergence in mathematics 8 See also 9 Bibliography 10 External links My two specific suggestions are:
1) separate emergent culture and emergent engineering. They seem quite distinct to me.
2) Create a section entitled Examples of Emergence then put many of the existing sections, as 'three-equal-sign' (===sect-name===) sub-sections. Thus, the revised outline might look something like this:

1 Emergent properties

2 Emergent concept

3 Examples of Emergent Outcomes

..Emergent structures in nature

..Emergence in human culture

..Emergence in physics

..Emergence in engineering

..- Cybernetic Emergence

..Emergence in pure mathematics

..Emergence in games

4 See also

5 Bibliography

6 External links

What do others think? N2e 18:57, 12 October 2006 (UTC)

I've been watching this article for a while, and I haven't noticed all that much activity on it. I suggest you go ahead and improve the article. It doesn't sound like you intend to delete anything, or add anything unreasonable, or make changes that don't make sense. So go for it. Be bold. WAS 4.250 19:34, 12 October 2006 (UTC)

Reorganization is a good idea. I agree that the current division needs a revision. Yet "Examples of Emergent Outcomes" is not a good title, since "outcomes" are not essential for the concept in general. They are more important for unintended consequences. A clarification of the relationship between emergence and unintended consequences would be nice.

A suitable classification or division of the different examples is not easy, since it is a very broad and general concept, and there is of course no strict agreement on exactly what the term means. One scientist means this, the other means that, one uses the word in this way, the other in a completely different way, and everyone is convinced that his way is the only right way. Everyone has his own world-views and preferences. I prefer the classification we use in the DCS-Wiki, but it might be too specific here.

However, "cybernetic emergence" should be deleted, as a single subsection it is superfluous, and the term itself is neither common, nor a suitable title for a section. The section "emergent concept" does not cite suitable references or sources, it seems to be a philosophical notion. Emergence in physics could be merged with emergent structures in nature, since physics decribes nature. Perhaps a section "Nature" with two subsections "non-living systems (physical)" and "living systems (biological)" would be suitable. --JFromm 16:50, 18 October 2006 (UTC)

Thanks for the input. Sounds like no consensus yet forming on this. Re cybernetic emergence, no problem if that term is not the best. I do believe that the example quoted in that section, where a sort of macro-behaviour can emerge amongst a large number of autonomous cybernetic devices, is a useful example of the breadth of emergent behaviour. So, change the term, but I recommend keeping the concept. And I agree that the entire article should be reorganized. N2e 06:00, 19 October 2006 (UTC)

I think it looks better now after a bit reconfiguration. The content is unchanged, only the sections are rearranged. --JFromm 16:37, 20 October 2006 (UTC)

Color example

Just a minor detail: I would say that atoms "absorb and reflect light", rather than "absorb and emit". Emission in this context would be an active act of generating the light which an atom doesn't do. The light (ie its different wavelengths) reflects off and is absobed by the atom, the result of which is then perceived as a specific color.

Atoms certainly can and do emit light! Every time an electron changes orbitals in an atom, it must absorb or emit a photon. You might change the article to say "absorb, emit, or reflect light", but I think it is fine as is. Grokmoo 20:00, 29 November 2006 (UTC)