|
|
Apologizes if this conversation has already continued elsewhere.
I don't know about Wimsatt, but I believe R. Rosen made similar points; i.e., an emergent property is essentially something that cannot be simulated. I disagree of course, but that's what he said, IIRC.
-Ted On Sun, Nov 8, 2009 at 4:39 PM, Nicholas Thompson <[hidden email]> wrote:
OK, so to you, I gather, emergence is just "nominal". The words at one level of abstraction do not apply readily to the entities at a lower level, and so emergence ... uh ... emerges.
So, you want to say, pressure is an emergent property of an arrangement of gas molecules in which they are all piled up on one place and all absent from another place nearby.
I hope others at this point will see the error of my ways, but.... I think I have to sign on to that.
Does anybody else remember that part of Wimsatt that I have coded in my brain as "nothing that is emergent can be studied by scientists". I know that is wrong, but what is it that he said that was like that. Roger??? Or are you in Chicago. Can you THINK in Chicago???
Nick
Nicholas S. Thompson
Emeritus Professor of Psychology and Ethology,
----- Original Message -----
To: [hidden email]
Sent: 11/8/2009 1:25:41 PM
Subject: Re: [FRIAM] Crutchfield 's "Is anything ever new?"
NST===> Now we can PUSH. <===nst
Speaking of pushing, the thing about pressure is that it seems like one of the standard examples of emergence. It's emergence at a relatively simple level--what I call static emergence--but emergence nevertheless. You did point out that it depends on the gas being inside the container. But is that enough for it to pass the test of being dependent on arrangement?
Pressure is a property of what? Not of the gas and the container but only of the gas in a confined area. The mechanism of confinement is presumably not relevant. According to Eric Weisstein, pressure is force per unit area, which I think is the standard definition. So what does it mean to ask whether pressure is an emergent property? What does it mean to say that pressure is a property at all? Presumably it means that it is a property of whatever is applying a force to an area.
The only reason we can talk about force per unit area is that we have statistically eliminated/aggregated the effects of the individual collisions of the gas molecules with the surface. So pressure would seem to be an emergent property (intuitively understood) of a gas that is understood scientifically by factoring out(!) the specific arrangement of elements. That seems to be where the contradiction arises. It's exactly the opposite of requiring that emergence depend (explicitly) on arrangements in time or space.
What about other static properties? What about hardness? Steven Weinberg notes that hardness is an emergent property of diamonds -- and that they have that property because of the way the carbon molecules are arranged. But the reason an object has a property is different from the property itself. If one wants to talk only about a property itself, a property that (let's assume) could be implemented in a number of different ways, then the particular way hardness is implemented in diamonds need not be fundamental to the property of hardness. So again, it seems that the property of hardness (as distinct from the mechanism of its implementation) factors out time and space.
That, of course, is my position. A property is emergent if it is a property of a level of abstraction. It makes no difference (according to me) how that property is implemented. Presumably a level of abstraction could be implemented in any number of ways.
To take an example from my own field, many devices are Turing complete, meaning that they are capable of computing any computable function. Being Turing complete is a property. Is it emergent? Not according to the requirement that it depends on an arrangement of time and space. There are numerous different ways of building Turing complete devices. Perhaps each one depends on a particular arrangement of component elements. But no particular arrangement is essential to being Turing complete.
-- Russ
On Sun, Nov 8, 2009 at 10:43 AM, Nicholas Thompson <[hidden email]> wrote:
This is good Russ; we are getting somewhere. we have locked horns. Now we can PUSH.
Please see below.
N
Nicholas S. Thompson
Emeritus Professor of Psychology and Ethology,
----- Original Message -----
Sent: 11/8/2009 11:06:43 AM
Subject: Re: [FRIAM] Crutchfield 's "Is anything ever new?"
OK. Then we have to ask what we mean by a property.
NST===>Agreed. And to be honest, I dont have an ambient definition. So I guess I will have to accept yours. <===nst
One standard definition is that a property is a predicate, i.e., a function mapping a thing to True or False. It then seems that as I mentioned before, according to the proposed definition, non-emergent properties are those that have factored out dependencies on the arrangement or timing of the elements that make up the whole [thing] -- a standard example being mass.
So is that approach as good, i.e., to define the property (of properties) of being non-emergent. A property is non-emergent if it has factored out any dependencies on arrangement ... .
NST===>Agreed again. You lay out pefectly Wimsatt's definition of the complement of emergence ... aggregativity. <===nst
I put it that way because if we suppose that we are talking about "reality," which is at one basic level an arrangement of stuff in time and space, then when we do science or make other abstractions about the world, we sometime factor out features of the world that we find can be ignored for certain purposes. When we can make such abstractions and they turn out to be useful, we have made a scientific advance.
NST===>Again. Perfect Wimsatt. I am completely on board. <===nst
So I'm not criticizing doing this: mass in Newtonian physics has worked quite well. But doesn't this imply that according to the proposed definition, emergent properties are those that haven't (completely) factored out that aspect of reality?
NST===> Exactly. Now we get to the hard part. The part of Wimsatt's article where he might be saying that one can only do science on non-emergent properties. I am so old and forgetful that I will have to go back and look, but this was a part of the paper we didnt discuss at length in the seminar and I may not have "gotten" it. <===nst
Would Newtonian momentum be non-emergent because it depends on directionality (arrangement) and speed (which depends on time)?
NST===> Oh Gosh. I need to specify "arrangement" dont I? Ugh. In other words, just saying that all the parts are in one place is not an adequate use of "arrangement" for W-emergence to work. I have to talk about relative arrangement ... internal arrangement.... . I wonder what trouble THAT gets me into. <===nst
How about statistical properties, which factor out arrangement and time? Wouldn't the proposed definition say that pressure, for example, is non-emergent because it doesn't depend on arrangement or time?
NST===> I think any "symmetrical" arrangement (in that mindblowing stupid way that physicists abuse that word) of particles could not be the basis of emergence. So, cautiously, I think I would have to agree that pressure is not an emergent property of the gas, though,, of course it is an emergent property of gas+vessel. I guess. <===nst
NST===>Have I walked into a crucial contradiction, here? I feel the vultures circling overhead. <===nst
-- Russ A
On Sun, Nov 8, 2009 at 9:28 AM, Nicholas Thompson <[hidden email]> wrote:
Russ,
I said: So, I suggest we define an emergent property as one that depends on the arrangement or timing of the elements that make up the whole.
You said: According to the proposed definition of emergence that page--with its component letters--is emergent.
I say: But a page is not a PROPERTY. I am prepared to stipulate that under "my" defintion (Wimsatt's definition) a great many boring properties are emergent, but you are taking it too far. It does have to be a property and the property cannot be a restatement of the arrangement or ordering of the elements that is the occasion for the emergence. And I do stipulate that using W.'s definition I will later have to shoulder the burden of identifying which sorts of emergence are interesting.
Nick
Nicholas S. Thompson
Emeritus Professor of Psychology and Ethology,
----- Original Message -----
To: [hidden email]; [hidden email]
Sent: 11/8/2009 12:18:51 AM
Subject: Re: [FRIAM] Crutchfield 's "Is anything ever new?"
I don't think Eric's point goes very far. A page with letters on it has letters as elements. According to the proposed definition of emergence that page--with its component letters--is emergent. Also, it doesn't matter whether the letters are arranged to have a meaning--in English or any other language. Any random collection of letters is emergent according to the proposed definition. It doesn't seem particularly useful to me to say that.
-- Russ A
On Sat, Nov 7, 2009 at 6:18 PM, Nicholas Thompson <[hidden email]> wrote:
Yeah. Like Eric Said!
n
Nicholas S. Thompson
Emeritus Professor of Psychology and Ethology,
----- Original Message -----
To: [hidden email]
Sent: 11/7/2009 7:09:10 PM
Subject: Re: [FRIAM] Crutchfield 's "Is anything ever new?"
In an attempt to defend Nick's definition (though I liked it better when he offered the categories of definitions than when he tried to pick one as proper): I suspect the statement "the series of letters in this sentence depends on the series of letters in this sentence", doesn't work, because the letters are not an element of the letters. That is, the definition offered requires a statement about something and its elements, not something and itself. Thus, you would need to say that "the sentence depends on the series of letters in the sentence", which is not terribly interesting to me, but is certainly not a tautology or otherwise trivial.
The only way I can see for you to try to argue back is to place especial emphasis on "the series" is the first phrase and "the letters" in the second. However, as soon as you are willing to consider "the sequence" as a real entity existing on a higher level, you are admitting emergence, and so the claim is not trivial (i.e., you have implicitly admitted from the start that "a sequence" is a variety of emergent).
Eric On Sat, Nov 7, 2009 08:35 PM, Russ Abbott <[hidden email]> wrote:
If something satisfies a definition (X is emergent if the elements of x are dependent on their arrangement ...) then what sense does it make to say that the definition doesn't apply to if it's satisfied trivially? It's still satisfied.
(Of course the dirt in your garden is also emergent under this criterion.)
It would seem that every property that doesn't abstract away arrangement and time becomes emergent. The mass of an aggregation is not emergent because mass abstracts away arrangement and time.
-- Russ A
On Sat, Nov 7, 2009 at 5:19 PM, Nicholas Thompson <nickthompson@...> wrote:
The arrangement is indeed dependent on the arrangement, but that's a tautology, and I dont think I am committed to tautologies because of my allegiance to Wimsattian emergence. The MEANING of the words of this sentence is indeed emergent since it is dependent on the arrangement of the letters. I am happy with the implication that a great many properties become emergent under the defintion. Contra Searle and a bunch of other people, I think emergence is as common as dirt .... well perhaps not quite that common.
N
Nicholas S. Thompson
Emeritus Professor of Psychology and Ethology,
----- Original Message -----
Sent: 11/7/2009 5:54:45 PM
Subject: Re: [FRIAM] Crutchfield 's "Is anything ever new?"
One problem with Nick's proposed definition is that it will label as emergent all sorts of uninteresting properties -- such as the sequence of characters in this message. I'm not talking about the semantics of the message or anything at all interesting, just the sequence of characters. That satisfies both of Nick's criteria.
So does the arrangement of molecules of air in your kitchen at exactly 3:00pm tomorrow. That satisfies the criterion of depending on the arrangement of elements.
-- Russ Abbott
_____________________________________________
Professor, Computer Science
California State University, Los Angeles
Cell phone: 310-621-3805
o Check out my blog at http://russabbott.blogspot.com/
On Sat, Nov 7, 2009 at 10:25 AM, Nicholas Thompson <nickthompson@...> wrote:
I agree that defitions, like everything else in science, should be heuristic.
So, I suggest we define an emergent property as one that depends on the arrangement or timing of the elements that make up the whole. In so defining emergence, we are led to ask, in every case of putative emergence, what is the particular arrangment or timing of presentation of the parts that makes this property possible.
Now, the tricky bit comes when we SUSPECT that a property is emergent but have not yet discovered (or think perhaps we may NEVER discover) the arrangments of parts that makes it possible. I gather that some properties of CA's fall into that category. Not sure what to do. We could, I suppose, define a loose category of "putative emergence" using surprise as a criterion, but reserve the term "emergent" itself for a property whose dependence on arrangment and/or timeing has been demonstrated.
It's heuristic because it leads to research.
Nick
Nick
Nicholas S. Thompson
Emeritus Professor of Psychology and Ethology,
----- Original Message -----
Sent: 11/7/2009 10:02:05 AM
Subject: Re: [FRIAM] Crutchfield 's "Is anything ever new?"
100%, complete, total unequivocal agreement w/Glen.
--Doug
--
Doug Roberts
droberts@...
doug@...
505-455-7333 - Office
505-670-8195 - Cell
On Fri, Nov 6, 2009 at 11:15 AM, glen e. p. ropella <gepr@...> wrote:
Thus spake Nicholas Thompson circa 11/05/2009 05:04 PM:
> I think your rejection of emergence applies only to (2) above.... and
> possibly (4), if we understand "no way" to mean "no way we have thought of
> yet". But I bet you disagree.
Naaa. I don't really disagree. I said I TEND to think that emergence
is fictitious. Until I see a definition or construction of it that I
can _use_ to get my work done, it's a useless concept, regardless of
whether it exists or not. I don't frankly care if it exists. What
matters is whether it can be used for some purpose (other than passing
the time arguing with bright people on e-mail lists ;-).
--
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
Eric Charles 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
============================================================
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
|
Locked
