Emergence and explanation
Locked
12
12
 
|
In this post, Eliezer Yudkowsky argues that it is
futile to use the word "emergence". Do you agree? http://lesswrong.com/lw/iv/the_futility_of_emergence/ -J. ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
Thus spake Jochen Fromm circa 09-07-01 12:02 AM:
> In this post, Eliezer Yudkowsky argues that it is futile to use the word > "emergence". Do you agree? > http://lesswrong.com/lw/iv/the_futility_of_emergence/ Yes, I agree. The only caveat is that the term "emergence" seems to mean something to a large swath of people. Hence, if you have a sense for that meaning and you're trying to invoke that meaning in the audience minds, then it's reasonable to use the term. But _merely_ for invoking that meaning in their minds, not to _explain_ anything but to manipulate them into thinking the way you want them to be thinking. Hence, I regard "emergence" as a purely manipulative term... a bit of rhetorical trickery intended to hypnotize the audience. Like all buzzwords, it's useful in some circumstances. So, it's not futile at all. But it should be used appropriately. -- glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
|
|
In reply to this post by Jochen Fromm-4
I am near to the end of plowing through the collection of articles on emergence (Editors' names slip me) and I think any quick embracing or dismissal of the concept is probably premature. I am hoping to get a bunch of people together in Santa Fe to read the collection and come to some common understanding of the disagreement. Most of the authors in the collection agree that emergence is no problem: the problem is that half of them do so because they believe that emergence is obviously occurs and the other half believes that it is just another case of philosophical word magic. I would love to develop a shared way of thinking about this, because I suspect that the impediment to a shared language is merely ideological and could be surmounted. .
My present candidate is Wimsatt's view which is to say that an entity has emergent properties if it has properties that depend upon the organization of its parts, rather than solely on the nature of the parts themselves. So a triangular frame has emergent properties not shared, for instance, with a parallelogram frame (other than having three members).. Nick -----Original Message----- >From: "glen e. p. ropella" <[hidden email]> >Sent: Jul 7, 2009 4:11 PM >To: The Friday Morning Applied Complexity Coffee Group <[hidden email]> >Subject: Re: [FRIAM] Emergence and explanation > >Thus spake Jochen Fromm circa 09-07-01 12:02 AM: >> In this post, Eliezer Yudkowsky argues that it is futile to use the word >> "emergence". Do you agree? >> http://lesswrong.com/lw/iv/the_futility_of_emergence/ > >Yes, I agree. The only caveat is that the term "emergence" seems to >mean something to a large swath of people. Hence, if you have a sense >for that meaning and you're trying to invoke that meaning in the >audience minds, then it's reasonable to use the term. But _merely_ for >invoking that meaning in their minds, not to _explain_ anything but to >manipulate them into thinking the way you want them to be thinking. > >Hence, I regard "emergence" as a purely manipulative term... a bit of >rhetorical trickery intended to hypnotize the audience. Like all >buzzwords, it's useful in some circumstances. So, it's not futile at >all. But it should be used appropriately. > >-- >glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com > > >============================================================ >FRIAM Applied Complexity Group listserv >Meets Fridays 9a-11:30 at cafe at St. John's College >lectures, archives, unsubscribe, maps at http://www.friam.org PS --Please if using the address [hidden email] to reply, cc your message to [hidden email]. Thanks. ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
Thus spake Nick Thompson circa 09-07-07 01:28 PM:
> My present candidate is Wimsatt's view which is to say that an entity > has emergent properties if it has properties that depend upon the > organization of its parts, rather than solely on the nature of the > parts themselves. So a triangular frame has emergent properties not > shared, for instance, with a parallelogram frame (other than having > three members).. But, as Yudkowsky said in that post that Jochen forwarded, isn't this true of _everything_? Can you name any system where _every_ property of the system is based solely on the nature of its components and not on its organization? More generally, is any property _not_ an emergent property? Sure, there may be _types_ of organization, as in the case of a triangle vs. a parallelogram; but if that's the case, why not just talk about types of organization instead of using the magical term "emergence"? -- glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
|
|
In reply to this post by Jochen Fromm-4
Well, If you read a textbook the physics text that we did a bit this year you find out that system is defined as anything you happen to be looking at. So any bunch of stuff would be a system. But any bunch of stuff does not display emergent properties. So on that account you would be wrong.
But I HATE that use of the term system. I would much prefer that we reserve the word system for an organization of some sort where the properties of the parts are interdependent. In which case, every system has emergent properties by definition and I absolutely agree with you. I am not on my own computer here, so wont say more now and may wish that I had said less. All the best, Nick -----Original Message----- >From: "glen e. p. ropella" <[hidden email]> >Sent: Jul 7, 2009 5:11 PM >To: The Friday Morning Applied Complexity Coffee Group <[hidden email]> >Subject: Re: [FRIAM] Emergence and explanation > >Thus spake Nick Thompson circa 09-07-07 01:28 PM: >> My present candidate is Wimsatt's view which is to say that an entity >> has emergent properties if it has properties that depend upon the >> organization of its parts, rather than solely on the nature of the >> parts themselves. So a triangular frame has emergent properties not >> shared, for instance, with a parallelogram frame (other than having >> three members).. > >But, as Yudkowsky said in that post that Jochen forwarded, isn't this >true of _everything_? Can you name any system where _every_ property of >the system is based solely on the nature of its components and not on >its organization? > >More generally, is any property _not_ an emergent property? > >Sure, there may be _types_ of organization, as in the case of a triangle >vs. a parallelogram; but if that's the case, why not just talk about >types of organization instead of using the magical term "emergence"? > >-- >glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com > > >============================================================ >FRIAM Applied Complexity Group listserv >Meets Fridays 9a-11:30 at cafe at St. John's College >lectures, archives, unsubscribe, maps at http://www.friam.org PS --Please if using the address [hidden email] to reply, cc your message to [hidden email]. Thanks. ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
|
|
In reply to this post by Jochen Fromm-4
Drafted most of this when the link first
came out. Finished now as interest might be returning: I completely agree with the criticism of how many people use the term "emergence", but I think the term itself is still useful. Yudkowsky's criticism could have been easily reworded as a guide to using the term properly. Of course, then it wouldn't have been as good of a read. As for the value of the term, a favorite quote of mine regarding emergence is: Coral reefs in the last analysis consist of positive and negative ions, but the biologist, geographer, or sea-captain would miss his point if he conceived them in any such terms. (Holt, 1915, p. 161) Whatever it is about coral reefs that would lead a sea captain to not think about it as a collection of ions... that is an emergent phenomenon. The biologist, the geographer and the sea captain are each interested in different properties that are not present in any given ion, but are present in the configuration of ions that make up a reef. The question of how collections of ions come to gain those properties needs to be treated as a serious scientific question. Otherwise it IS just another word for magic. Of course, as Glen mentions, everything is "emergent" in this sense. However, not everything emerges in the same way, and the thing to be investigated is how different things emerge. The properties that the different people are interested in will be products of different aspects of the configuration. As an example of the word magic: In my experience, people who say things like "Intelligence is an emergent product of neurons firing" are typically using the term "emergent" as an excuse for not doing the hard work, rather than as a marker for where the hard work needs to be done. Their assertion should mean something like, "Models that explain the origins of intelligence can be produced using only configurations of interconnected neurons, and we are interested in how that particular kind of interconnection works." However, in practice it usually means something more like, "We don't understand intelligence, it seems bloody mysterious, and we would really rather not talk about it any more." Eric On Wed, Jul 1, 2009 03:02 AM, "Jochen Fromm" <[hidden email]> wrote: In this post, Eliezer Yudkowsky argues that it is ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
In reply to this post by Nick Thompson
Thus spake Nick Thompson circa 09-07-07 06:08 PM:
> Well, If you read a textbook the physics text that we did a bit this > year you find out that system is defined as anything you happen to be > looking at. So any bunch of stuff would be a system. But any bunch > of stuff does not display emergent properties. So on that account you > would be wrong. Well, since my post consisted of questions, I could hardly be wrong. ;-) The question was: Is there any identifiable property of a system that is NOT an emergent property, regardless of how one defines "system"? If anyone knows of one, please name it! Yes, I am _partly_ asking so that I can subsequently analyze that example and demonstrate that whatever example is provided, it can be thought of as "emergent". I'm also genuinely interested in the examples list members might assert are non-emergent properties. Honestly, I can't think of any. > -----Original Message----- >> From: "glen e. p. ropella" <[hidden email]> >> >> [...] >> >> But, as Yudkowsky said in that post that Jochen forwarded, isn't >> this true of _everything_? Can you name any system where _every_ >> property of the system is based solely on the nature of its >> components and not on its organization? >> >> More generally, is any property _not_ an emergent property? >> >> Sure, there may be _types_ of organization, as in the case of a >> triangle vs. a parallelogram; but if that's the case, why not just >> talk about types of organization instead of using the magical term >> "emergence"? -- glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
On Wed, Jul 08, 2009 at 10:16:55AM -0700, glen e. p. ropella wrote:
> > Well, since my post consisted of questions, I could hardly be wrong. ;-) > > The question was: Is there any identifiable property of a system that is > NOT an emergent property, regardless of how one defines "system"? If > anyone knows of one, please name it! Absolutely! The positions of the particles in a Newtonian n-body system are not emergent. Of course there are other properties of these systems that are emergent, but position & momenta of the particles are not amongst them, being part of the basic vocabulary of the model. > > Yes, I am _partly_ asking so that I can subsequently analyze that > example and demonstrate that whatever example is provided, it can be > thought of as "emergent". I'm also genuinely interested in the examples > list members might assert are non-emergent properties. Honestly, I > can't think of any. > -- ---------------------------------------------------------------------------- Prof Russell Standish Phone 0425 253119 (mobile) Mathematics UNSW SYDNEY 2052 [hidden email] Australia http://www.hpcoders.com.au ---------------------------------------------------------------------------- ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
|
|
On Wed, Jul 8, 2009 at 8:33 PM, russell standish<[hidden email]> wrote:
> On Wed, Jul 08, 2009 at 10:16:55AM -0700, glen e. p. ropella wrote: >> The question was: Is there any identifiable property of a system that is >> NOT an emergent property, regardless of how one defines "system"? Â If >> anyone knows of one, please name it! > > Absolutely! The positions of the particles in a Newtonian n-body system > are not emergent. Of course there are other properties of these > systems that are emergent, but position & momenta of the particles are > not amongst them, being part of the basic vocabulary of the model. OK, so aren't the positions and velocities of the particles a consequence of the forces affecting the particles? Is saying something is a consequence the same as saying it is an emergent property? Or is this concequence too well defined, which brings us back to "emergent" means "poorly understood"? I'm not sure why I'm even playing this game, since I don't think its helpful to say that everything is an emergent property of something... because of course it is... because everything in the universe is made of math. ~~James ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
|
|
In reply to this post by Jochen Fromm-4
James,
I tried this earlier, and I dont know what you thought: If a property of a whole is sensitive not only to the properties of the its but to their arrangement, or order of presentation, etc., then that property is an emergent one. Some writers have tried to put us off the scent by calling this weak emergence, but it IS the difference we are interested in, isnt it? I have never been able to identify any examples of "strong" emergence that didnt smack of mysticism. There is one aspect of this definition that makes me uneasy: it is an EXPLANATORY definition, like defining adaptation as "whatever natural selection produces". This means that we have to have a [reductive!] explanation for a property in place before we can say whether it is an emergent property or not. In the case of natural selection, reliance on explanatory definition has led to a lot of uncritical circular reasoning. Also, it depends on a clear understanding of what it is to be a "property of a part". I think to be a property of a part means that you cannot mention any other part in the description of that part. So, "being on the left" or "being added to the pile first" are not properly properties of parts. The definition makes emergence VERY common, of course, particularly in the world of computers where the same parts elements get moved around to produce many different kinds of outcomes and similar outcomes can be produced by different arrangements of the parts (multiple realizability). But I would rather have "emergence" banal than mystic. Do you have the EMERGENCE (Bedau and Phillips, MIT press, 2008) collection? Shall we do an online seminar on it???? We could agree to read an essay a week and comment on it. I could set up a WIKI. Shared phone line. Gotta go. Nick Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University ([hidden email]) http://home.earthlink.net/~nickthompson/naturaldesigns/ > [Original Message] > From: James Steiner <[hidden email]> > To: The Friday Morning Applied Complexity Coffee Group <[hidden email]> > Date: 7/9/2009 6:27:33 AM > Subject: Re: [FRIAM] Emergence and explanation > > On Wed, Jul 8, 2009 at 8:33 PM, russell standish<[hidden email]> wrote: > > On Wed, Jul 08, 2009 at 10:16:55AM -0700, glen e. p. ropella wrote: > >> The question was: Is there any identifiable property of a system that is > >> NOT an emergent property, regardless of how one defines "system"? Â If > >> anyone knows of one, please name it! > > > > Absolutely! The positions of the particles in a Newtonian n-body system > > are not emergent. Of course there are other properties of these > > systems that are emergent, but position & momenta of the particles are > > not amongst them, being part of the basic vocabulary of the model. > > OK, so aren't the positions and velocities of the particles a > consequence of the forces affecting the particles? Is saying something > is a consequence the same as saying it is an emergent property? Or is > this concequence too well defined, which brings us back to "emergent" > means "poorly understood"? > > I'm not sure why I'm even playing this game, since I don't think its > helpful to say that everything is an emergent property of something... > because of course it is... because everything in the universe is made > of math. > > ~~James > > ============================================================ > FRIAM Applied Complexity Group listserv > Meets Fridays 9a-11:30 at cafe at St. John's College > lectures, archives, unsubscribe, maps at http://www.friam.org ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
In reply to this post by James Steiner
Actually-
Everything in the world is not made of math, it is made of relationships, which math as a language describes better than anything else we have (commonly agreed upon). Not much difference, perhaps, however emergence or whatever it is that we reference is also a property of the relationship between things. Relate - 'to tell back, to give back' Tory On Jul 9, 2009, at 6:25 AM, James Steiner wrote: > On Wed, Jul 8, 2009 at 8:33 PM, russell standish<[hidden email] > > wrote: >> On Wed, Jul 08, 2009 at 10:16:55AM -0700, glen e. p. ropella wrote: >>> The question was: Is there any identifiable property of a system >>> that is >>> NOT an emergent property, regardless of how one defines "system"? >>> If >>> anyone knows of one, please name it! >> >> Absolutely! The positions of the particles in a Newtonian n-body >> system >> are not emergent. Of course there are other properties of these >> systems that are emergent, but position & momenta of the particles >> are >> not amongst them, being part of the basic vocabulary of the model. > > OK, so aren't the positions and velocities of the particles a > consequence of the forces affecting the particles? Is saying something > is a consequence the same as saying it is an emergent property? Or is > this concequence too well defined, which brings us back to "emergent" > means "poorly understood"? > > I'm not sure why I'm even playing this game, since I don't think its > helpful to say that everything is an emergent property of something... > because of course it is... because everything in the universe is made > of math. > > ~~James > > ============================================================ > FRIAM Applied Complexity Group listserv > Meets Fridays 9a-11:30 at cafe at St. John's College > lectures, archives, unsubscribe, maps at http://www.friam.org > ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
In reply to this post by Russell Standish
Thus spake russell standish circa 07/08/2009 05:33 PM:
> On Wed, Jul 08, 2009 at 10:16:55AM -0700, glen e. p. ropella wrote: >> Well, since my post consisted of questions, I could hardly be wrong. ;-) >> >> The question was: Is there any identifiable property of a system that is >> NOT an emergent property, regardless of how one defines "system"? If >> anyone knows of one, please name it! > > Absolutely! The positions of the particles in a Newtonian n-body system > are not emergent. Of course there are other properties of these > systems that are emergent, but position & momenta of the particles are > not amongst them, being part of the basic vocabulary of the model. Excellent! Thanks Russell. However, I claim that the positions and momenta (note the plural) of the individual components are not properties of the _system_. Those are properties of the individual components. A systemic property related to those component properties might be a centroid or cumulative (averaged, summed, etc.) momentum for the system as a whole. Of course, the position or momentum of an individual particle is a systemic property of the system that constitutes that single particle (a system of quarks, say). The question then becomes, is a centroid or cumulative measure of a system of particles "emergent"? Or are the position and momentum of a system of quarks "emergent"? -- glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
In reply to this post by Nick Thompson
Thus spake Nicholas Thompson circa 09-07-09 06:52 AM:
> Also, it depends on a clear understanding of what it is to be a > "property of a part". I think to be a property of a part means that > you cannot mention any other part in the description of that part. Excellent! This demonstrates quite well why it is incoherent to say that a systemic property is non-emergent. It is logically impossible to describe a _part_ of a system without describing the context or environment into which that part fits, namely the other parts of the system. Further, to describe any _unit_... any object with a boundary around it, you must distinguish that unit from the ambience around it. I.e. you can't describe the object without at least partially describing the NOT-object. So, the root of the incoherence of "emergent" lies in an inability to define a closure. (Unleash the Rosenites! ;-) > So, "being on the left" or "being added to the pile first" are not > properly properties of parts. And neither is position or momentum because they both have to be defined _relative_ to something, trivially to an arbitrary vector space origin, non-trivially to other particles. Unless you're treating the particle as a system, itself. And then position and momentum are emergent properties of the sub-particle components. So, either way, they are a result of the systems organization and the interaction of their components. Emergence is a trivial (but not entirely useless) word except in the sense of emergency: "A serious situation or occurrence that happens unexpectedly and demands immediate action", which boils down to "poorly understood" or, at least, unpredictable. -- glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
In reply to this post by glen e. p. ropella-2
Hm.
What about properties that show up (a synonym for 'emerge') in one system but also show up in other systems. Many things reposition when in combination with other things/dynamics. Does this mean that all those repositionings are emergent? Or just a physical law that applies across a range of systems? Tory On Jul 9, 2009, at 10:11 AM, glen e. p. ropella wrote: > Thus spake russell standish circa 07/08/2009 05:33 PM: >> On Wed, Jul 08, 2009 at 10:16:55AM -0700, glen e. p. ropella wrote: >>> Well, since my post consisted of questions, I could hardly be >>> wrong. ;-) >>> >>> The question was: Is there any identifiable property of a system >>> that is >>> NOT an emergent property, regardless of how one defines "system"? >>> If >>> anyone knows of one, please name it! >> >> Absolutely! The positions of the particles in a Newtonian n-body >> system >> are not emergent. Of course there are other properties of these >> systems that are emergent, but position & momenta of the particles >> are >> not amongst them, being part of the basic vocabulary of the model. > > Excellent! Thanks Russell. > > However, I claim that the positions and momenta (note the plural) of > the > individual components are not properties of the _system_. Those are > properties of the individual components. A systemic property > related to > those component properties might be a centroid or cumulative > (averaged, > summed, etc.) momentum for the system as a whole. > > Of course, the position or momentum of an individual particle is a > systemic property of the system that constitutes that single > particle (a > system of quarks, say). > > The question then becomes, is a centroid or cumulative measure of a > system of particles "emergent"? Or are the position and momentum of a > system of quarks "emergent"? > > -- > glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com > > > > ============================================================ > FRIAM Applied Complexity Group listserv > Meets Fridays 9a-11:30 at cafe at St. John's College > lectures, archives, unsubscribe, maps at http://www.friam.org > ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
Thus spake Victoria Hughes circa 09-07-09 11:21 AM:
> What about properties that show up (a synonym for 'emerge') in one > system but also show up in other systems. Many things reposition when in > combination with other things/dynamics. Does this mean that all those > repositionings are emergent? Or just a physical law that applies across > a range of systems? Heh, your language will easily draw us into into using "emergent" as if it has meaning. ;-) I don't like the word "property" for this context. So, I'll use "characteristic". Characteristics exhibited by measuring a particular system with a particular measure can show up when one applies the same measure to a different system. Hence, I'd say that two distinct systems that exhibit the same characteristic under the same measure are members of a set (class, family, range, group, whatever term you like) defined by that measure. Similarly, the two distinct systems can show _similar_ characteristics when two distinct measures are applied. In that ambiguous case, it's up to the observer to tease out whether the characteristics are similar because the systems are members of a well-defined set (i.e. a better measure can be defined that makes the set coherent) _or_ whether the characteristics are similar just because the measures are similar. But more to your point, it's reasonable that one component of a system is somehow canalizing under particular measures. Hence, if that component is present in several distinguishable systems, then applying those measures to all those systems shows the characteristics associated with that dominant component. As long as that component _requires_ a system in which to live (i.e. the measures can't be applied to the component, only to the system containing the component, which is true in all non-trivial situations), then the characteristics are still "emergent", even though we reasonably and abstractly reduce the cause to the component. -- glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
|
|
In reply to this post by Jochen Fromm-4
Glen,
Thanks for your thoughts, here. The correspondence got so untidy below that I thought best to restate the position as sharpened by your critique. I would say that, contra Atkins Physical Chemistry and many other experts, a system is more than just any old thing we happen to be talking about. To be a system, the thing we are talking about must be *organized*. Another way of saying this is that all systems have emergent properties: i.e., properties that arise from the *internal* arrangement or ordering of their parts. Arrangements or ordering of parts that are true of all of those parts are not internal. I don't think your critique is specific in any way to the problem of defining the emergent properties of things It relates generally to the definition of the primitive, "any old thing". Is a lens cloud an object? Is an ocean wave an object? A sand dune? An organism, for that matter? It comes up any time we try to justify the use of any concrete noun. I believe that "Thing" can be a primitive and we can still have a useful discussion of what the emergent properties of a thing are, or are not. An insistence on rigorous closure would be the end of all conversation, because we never would be able to meet the standard. Which, come to think of it, may be your point. You are arguing that the conversation we are trying to have is impossible? If you want to see the verbal train wreck that preceded this summary, look below, but I don't recommend it. Oh by the way: do you have the Bedau and Phillips book? Do you think it might be possible to have an international webSeminar on it? Colloquium? What would that look like? Nick Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University ([hidden email]) http://home.earthlink.net/~nickthompson/naturaldesigns/ > [Original Message] > From: glen e. p. ropella <[hidden email]> > To: The Friday Morning Applied Complexity Coffee Group <[hidden email]> > Date: 7/9/2009 11:06:37 AM > Subject: Re: [FRIAM] Emergence and explanation > > Thus spake Nicholas Thompson circa 09-07-09 06:52 AM: > > Also, it depends on a clear understanding of what it is to be a > > "property of a part". I think to be a property of a part means that > > you cannot mention any other part in the description of that part. > > Excellent! This demonstrates quite well why it is incoherent to say > that a systemic property is non-emergent. nst==>Right! But doesnt this mean we have to somehow get our colleagues to stop using word "system" to refer to "whatever we are talking about". <==nst It is logically impossible to > describe a _part_ of a system without describing the context or > environment into which that part fits, namely the other parts of the system. > > Further, to describe any _unit_... any object with a boundary around it, > you must distinguish that unit from the ambience around it. I.e. you > can't describe the object without at least partially describing the > NOT-object. So, the root of the incoherence of "emergent" lies in an > inability to define a closure. (Unleash the Rosenites! ;-) nst==> Sorry, I didn't follow this last bit. Let's assume that we can get the rest of the world to go along with our understanding of "system". Are you arguing here that every object has to be a system? I.e., every object has emergent properties? Hmm. I am wondering whether I agree with this..................................... I am tempted to argue that an object is a "pile of stuff that moves around together" Hmmmm. No, Too weak, because if a bulldozer comes along and picks up my pile of stuff and moves it to a new place, I may be tempted to claim that this object is no object at all because "accidental". Is a sand dune an object? Is a lens cloud an object? When I claimed that an emergent property of a whole is one that arises from the arrangement or ordering of the presentation of its parts I was not speaking of any arrangement that is true of all the parts. So, for instance, if all the parts are accelerating at the same speed relative to other wholes, this does not consist of an arrangement or ordering for the purposes of the definition. <===nst. > > > So, "being on the left" or "being added to the pile first" are not > > properly properties of parts. > to > And neither is position or momentum because they both have to be defined > _relative_ to something, trivially to an arbitrary vector space origin, > non-trivially to other particles. nst==><==nst e. Unless you're treating the particle > as a system, itself. And then position and momentum are emergent > properties of the sub-particle components. So, either way, they are a > result of the systems organization and the interaction of their components. > > Emergence is a trivial (but not entirely useless) word except in the > sense of emergency: "A serious situation or occurrence that happens > unexpectedly and demands immediate action", which boils down to "poorly > understood" or, at least, unpredictable. > > -- > glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com > > > ============================================================ > FRIAM Applied Complexity Group listserv > Meets Fridays 9a-11:30 at cafe at St. John's College > lectures, archives, unsubscribe, maps at http://www.friam.org ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
|
|
Trying to be more succinct than my last post. I am of the opinion
that:
"X is emergent.", as a strong claim, is surely word magic. Indeed, everything is emergent (minus the bare contents of some imaginary lowest level), so no insight is gained from the labeling - no explanation is added. In contrast, "X emerges from the combination of A, B, and C, when conditions P and Q are met," is a great thing to be able to say. It can answer crucial questions, can lead to great research programs, etc. Of course, one could drop the word entirely. In that case, one would just say "X is the combination of A, B, and C, when conditions P and Q are met." However.... uhm.... however... Damn. I thought I could come up with a good "however". Then I deleting seven things that seemed very convincing before I wrote them, but not afterwards. I can only hope someone on the list can come up with one, because I really like talking about emergence. Hoping for help, Eric On Thu, Jul 9, 2009 10:28 PM, "Nicholas Thompson" <[hidden email]> wrote: Eric CharlesGlen, Thanks for your thoughts, here. The correspondence got so untidy below that I thought best to restate the position as sharpened by your critique. I would say that, contra Atkins Physical Chemistry and many other experts, a system is more than just any old thing we happen to be talking about. To be a system, the thing we are talking about must be *organized*. Another way of saying this is that all systems have emergent properties: i.e., properties that arise from the *internal* arrangement or ordering of their parts. Arrangements or ordering of parts that are true of all of those parts are not internal. I don't think your critique is specific in any way to the problem of defining the emergent properties of things It relates generally to the definition of the primitive, "any old thing". Is a lens cloud an object? Is an ocean wave an object? A sand dune? An organism, for that matter? It comes up any time we try to justify the use of any concrete noun. I believe that "Thing" can be a primitive and we can still have a useful discussion of what the emergent properties of a thing are, or are not. An insistence on rigorous closure would be the end of all conversation, because we never would be able to meet the standard. Which, come to think of it, may be your point. You are arguing that the conversation we are trying to have is impossible? If you want to see the verbal train wreck that preceded this summary, look below, but I don't recommend it. Oh by the way: do you have the Bedau and Phillips book? Do you think it might be possible to have an international webSeminar on it? Colloquium? What would that look like? Nick Nicholas S. Thompson Emeritus Professor of Psychology and Ethology, Clark University ([hidden email]) http://home.earthlink.net/~nickthompson/naturaldesigns/ > [Original Message] > From: glen e. p. ropella <[hidden email]> > To: The Friday Morning Applied Complexity Coffee Group <[hidden email]> > Date: 7/9/2009 11:06:37 AM > Subject: Re: [FRIAM] Emergence and explanation > > Thus spake Nicholas Thompson circa 09-07-09 06:52 AM: > > Also, it depends on a clear understanding of what it is to be a > > "property of a part". I think to be a property of a part means that > > you cannot mention any other part in the description of that part. > > Excellent! This demonstrates quite well why it is incoherent to say > that a systemic property is non-emergent. nst==>Right! But doesnt this mean we have to somehow get our colleagues to stop using word "system" to refer to "whatever we are talking about". <==nst It is logically impossible to > describe a _part_ of a system without describing the context or > environment into which that part fits, namely the other parts of the system. > > Further, to describe any _unit_... any object with a boundary around it, > you must distinguish that unit from the ambience around it. I.e. you > can't describe the object without at least partially describing the > NOT-object. So, the root of the incoherence of "emergent" lies in an > inability to define a closure. (Unleash the Rosenites! ;-) nst==> Sorry, I didn't follow this last bit. Let's assume that we can get the rest of the world to go along with our understanding of "system". Are you arguing here that every object has to be a system? I.e., every object has emergent properties? Hmm. I am wondering whether I agree with this..................................... I am tempted to argue that an object is a "pile of stuff that moves around together" Hmmmm. No, Too weak, because if a bulldozer comes along and picks up my pile of stuff and moves it to a new place, I may be tempted to claim that this object is no object at all because "accidental". Is a sand dune an object? Is a lens cloud an object? When I claimed that an emergent property of a whole is one that arises from the arrangement or ordering of the presentation of its parts I was not speaking of any arrangement that is true of all the parts. So, for instance, if all the parts are accelerating at the same speed relative to other wholes, this does not consist of an arrangement or ordering for the purposes of the definition. <===nst. > > > So, "being on the left" or "being added to the pile first" are not > > properly properties of parts. > to > And neither is position or momentum because they both have to be defined > _relative_ to something, trivially to an arbitrary vector space origin, > non-trivially to other particles. nst==><==nst e. Unless you're treating the particle > as a system, itself. And then position and momentum are emergent > properties of the sub-particle components. So, either way, they are a > result of the systems organization and the interaction of their components. > > Emergence is a trivial (but not entirely useless) word except in the > sense of emergency: "A serious situation or occurrence that happens > unexpectedly and demands immediate action", which boils down to "poorly > understood" or, at least, unpredictable. > > -- > glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com > > > ============================================================ > FRIAM Applied Complexity Group listserv > Meets Fridays 9a-11:30 at cafe at St. John's College > lectures, archives, unsubscribe, maps at http://www.friam.org ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org Professional Student and Assistant Professor of Psychology Penn State University Altoona, PA 16601 ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
|
|
The first step in any conversation in any sane about emergence is, I think, to decide what sorts of things-that-we-refer-to-with-nouns are going to be candidates for the predicate "emergent". Objects? Or properties. (There are other choices but I forget them). Once we fix on "properties", I think our job becomes easier, because it seems to me self evident that almost system we might discuss has some emergent and some non-emergent properties. Breadiness is a wonderful example of an emergent property. It clearly depends on how you arrange the ingredients (you have to mix them) and the order of the processes you apply to the ingredients (best you should bake them after you mix them).
So, would agree with you if X is specified as a property in your formula.
N
Nicholas S. Thompson
Emeritus Professor of Psychology and Ethology,
Clark University ([hidden email])
============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
In reply to this post by glen e. p. ropella-2
<BASE href="file://C:\Users\Rikus\Documents\My Stationery\">
> The question was: Is there any identifiable property of a
> system that is NOT an emergent property, regardless of
> how one defines "system"? If anyone knows of one,
> please name it!
1. The number of components in a Newtonian N-body system.
In the study of complex systems, the number of components
is frequently a critical system property. It is,
however, entirely divorced from the properties of the components or
relationships among them.
2. A weaker one: the mass of a Newtonian N-body
system. While it depends on component masses, it is independent
of interaction among components.
Regards,
Rikus
-------------------------------------------------- From: "glen e. p. ropella" <[hidden email]> Sent: Wednesday, July 08, 2009 7:16 PM To: "The Friday Morning Applied Complexity Coffee Group" <[hidden email]> Subject: Re: [FRIAM] Emergence and explanation Thus spake Nick Thompson circa 09-07-07 06:08 PM: > Well, If you read a textbook the physics text that we did a bit this > year you find out that system is defined as anything you happen to be > looking at. So any bunch of stuff would be a system. But any bunch > of stuff does not display emergent properties. So on that account you > would be wrong. Well, since my post consisted of questions, I could hardly be wrong. ;-) The question was: Is there any identifiable property of a system that is NOT an emergent property, regardless of how one defines "system"? If anyone knows of one, please name it! Yes, I am _partly_ asking so that I can subsequently analyze that example and demonstrate that whatever example is provided, it can be thought of as "emergent". I'm also genuinely interested in the examples list members might assert are non-emergent properties. Honestly, I can't think of any. ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
Re: Emergence and explanation
|
|
In reply to this post by Nick Thompson
Thus spake Nicholas Thompson circa 07/09/2009 07:28 PM:
> I would say that, contra Atkins Physical Chemistry and many other experts, > a system is more than just any old thing we happen to be talking about. > To be a system, the thing we are talking about must be *organized*. I don't intend to argue about the meaning of the word "system". It's OK if you (and others) fixate on that; but it's not my intent. Having said that, let me try to get the point I actually want to make formulated in your terms. Organization is merely another characteristic of the goo/stuff we perceive (identify, carve out) of the ambience. Organization is no different from any other characteristic that obtains when we apply a predicate or measure to the ambience. To see this, think of any system you want, then think about the different predicates/measures/perspectives/hats you might adopt in order to get at the organization of the system. E.g. when analyzing a corporation, an accountant sees one organization and an entrepreneur sees something else. E.g. when analyzing a condensed gas a string theorist sees one organization and a traditional quantum physicist sees another. What you're doing when you say a system must be organized is this: You are _imputing_ an ontological property (organization) into a perceived slice of the goo/stuff. You're saying, assuming objectivity, that that particular bucket of goo/stuff has, regardless of observer, property X. > those parts are not internal. I don't think your critique is specific in > any way to the problem of defining the emergent properties of things It > relates generally to the definition of the primitive, "any old thing". Is a > lens cloud an object? Is an ocean wave an object? A sand dune? An > organism, for that matter? It comes up any time we try to justify the > use of any concrete noun. I believe that "Thing" can be a primitive and we > can still have a useful discussion of what the emergent properties of a > thing are, or are not. An insistence on rigorous closure would be the end > of all conversation, because we never would be able to meet the standard. > Which, come to think of it, may be your point. You are arguing that the > conversation we are trying to have is impossible? Excellent! But not quite what I'm arguing. I'm not arguing that such conversation is impossible. I'm arguing that it is _vague_ and cannot be completely resolved. And the reason it is vague is because it depends fundamentally on well-formulating 2 things: 1) the presumed objective system and 2) the measures used to observe the system. You guys have been leaving (2) off and assuming that "emergence" is or can be somehow independent of the perspective the observer takes. It can't. (Note that I'm not claiming emergence is a purely subjective thing. I'm claiming that the concept requires _both_ the objective and the subjective.) However, if you explicitly lay out your assumptions about the system (with no hidden secret meanings to the word "system") then lay out how you intend to _measure_ the system, then and only then can you have a (less vague) conversation about emergence. > Oh by the way: do you have the Bedau and Phillips book? Do you think it > might be possible to have an international webSeminar on it? Colloquium? > What would that look like? No, I don't have the book. It's possible to have a seminar about it; but unless it were about concrete things, particular systems, particular methods for measuring those systems, then I don't think its of much use. If you chose a family of systems (perhaps your statics example of triangular trusses) and a family of measures (perhaps robustness to harmonic oscillation), _then_ it would be interesting to talk about emergent characteristics of such systems as measured in that way. -- glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org |
«
Return to Friam
|
1 view|%1 views
| Free forum by Nabble | Edit this page |