Re: What is emergence?

[address@hidden]

> > > Consider a system described by 2 languages which we call 
"microscopic"
> > > and "macroscopic". Furthermore, we suppose that the system is
> > > completely described by the microscopic language - true in the 
case of
> > > an agent based simulation for example, arguably less true of a 
real
> > > world system, where the microscopic description is probably 
expressed
> > > in terms of an approximation like classical mechanics.
> > 
> > > If some concept or property in the macroscopic language is not
> > > contained within the microscopic description, then it is emergent. 
> > 

Well, now I think I have better understood what you mean with language.
In my opinion the expression "not contained" might be better expressed, 
couse it may bring us to some kind of misunderstanding.
>From your first posting you said that emergence is not only something
new, unpredictable etc., and I agree with this position.
So we can ask our self "Is emergence always a discovery ?", and we may
agree with "No".
Thus "not contained" may be confused with "New", and that's not what we
mean.

For example in the macroeconomic language there is the well known 
concept of "recession". The first time we *observe* recession we
*discover* it, recognize a *pattern* (occupation, inflation, etc.) and
name this pattern *recession*, adding a new term to the macro language
*not conteined* before.
Now there are a lot of efforts in studying "how recession emerges", 
but any time we percive recession (observe the defined pattern), 
we say that a new recession has emerged without adding nothing new
to our macrolanguage.

It is for this reason that I prefer to link the concept of emergence
to the observation of a pattern in a macro and micro world, each one
defined by the appropriate language.


With regard to the emergence at the microlevel (if definable), may be
I haven't choose the right examples. Now I will try with another one, 
may be it is not as broadly known as water and social behavior, but I
think it is quite common and easy to understand for thouse new to it.
This exaple is the study of the prisoners dilemma made by R. Axelrod
using genetic algorithms ("The evolution of strategies in the iterated
prisoner's dilemma" R. Axelrod, Genetic algorithms and simulated 
annealing ed. Lawrence Davis 1987, 32-41; you can find it in the book
"The complexity of cooperation: agent based models of competition and 
collaboration" R. Axelrod, Princeton 1997, ISBN:0-691-01567-8).

In this study any agent have its behavior coded in a gene: simply a 
vector where the index of the vector indicates the previos history
of the game, and the vector contains the next action to do given
the past history (Cooperate or Defect). So we can think of the gene
as a vector of C and D.
So the micro language describe the vector, how the agents interacts, and 
what happends if they cooperate or not (given agent a and b there are
payoffs defined for ab = CC, CD, DC, DD).
The micro language describe what is the vector and what does it means, 
the language cannot define where in the vector must be C or D, couse
it is the object of the experiment.
At the macro level the language define the mean score of the group
of agents (20 agents in the expriment).

Running the experiment the observer can see at the macro level that
the avarage score falls down (so defection emerges), and than rise up
( cooperation emerges, more precisely reciprocation).
But defection and cooperation are something defined within
the microlanguage as particular patterns of the gene.

This bring us to another question:
Is emergence a state or a process ?
I think it is a process.

The process terminates when we say that something "has emerged", meaning 
that the particular pattern has become *common* or *diffuse*.
But the main object of our studies on emergence is "how something
begins to emerge, under which conditions ..."
Economists study how can recession emerges, the article mentioned above
(among other things) try to explain why cooperation emerges after 
defection etc.

So at this point I propose this draft of definition:
"Emergence is a process that brings the observer of a system described
with a macro and micro language, to observe a common pattern. The 
pattern is described with micro and macro language and, if not previosly 
 defined by the languages, this process add new concepts at the 
languages used for the system."


With regard to postings about the interaction between agents and the 
environment, my humble opinion is that the environment is nothing but
an agent itself: it has a description, a behavior and so on.

Comments will be welcome, 

Matteo Piombo

Universita' di Padova (DIMEG)
Padova (PD)



                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================