« September 2007 | Main | November 2007 »

October 29, 2007

Interesting PBS Series

So, I don't have cable and watch alot of PBS. For the last few weeks I've happened upon this series called Guns, Germs & Steel which explores the economic disparities across the globe as an iterative function of the initial conditions encountered by early human settlers. To sum it up, its a critical look at the reasons why Europeans were the first to develop the technology necessary to exploit and enslave the rest of the world and how our current global distribution of wealth is very much the product of the "he who gets, gets more" model explored in our readings. Some interesting factoids from the series:

1) Worldwide there are only 14 species of large animals that have been successfully domesticated for agriculture.
2) The -1% growth (shrink) rate of the African economy can be attributed entirely to that continent's Malaria epidemic.

Prof. Jared Diamond, the evolutionary biologist behind the series, treads very close to the doctrines of chaos and complexity in many of his on-camera statements. Interesting stuff.

Check out the website: http://www.pbs.org/gunsgermssteel/about/jared.html

October 24, 2007

Wikis at UMN

Here is a website that allows groups at UMN to create wikis. This may or may not be too late for us, though you might also find it useful for the future.


October 18, 2007

For what it's worth . . .

strange attractors.jpg
Click on the image to view all of it

This is a sketch that I did during our group discussion in class last night. Our group ended up discussing two main topics: 1. In culture, how do you decide which agents compose a given complex adaptive system, and 2. In light of our members' unique interests in aspects of cultural change, how could we come to a common understanding of how cultural change can be effected.

This picture illustrates several aspects of complex adaptive systems (CAS). CAS's are often composed of agents that are sub-systems of the CAS in question. The CAS of interest may be a sub-system of another system. not all sub-systems are complex or adaptive, but all systems and sub-systems are open. That is, feedback loops exist both within a system and between a system and its environment (which undoubtedly includes other CAS's). Even systems that are static can be "consumed" by a CAS and become part of that system. This principle is represented by the Game of Life in which a complex entity will "eat" non-dynamic objects and leave others in its wake. The arrangement of the agents within each CAS and the links between them were drawn randomly, not to signify that in real CAS's these are necessarily random, but to demonstrate that arrangement, relationship, and dynamics are unique in different systems. However, self-similarity among different scales of CAS's indicate that they ar likely also fractal in nature. The things that this diagram doesn't really capture is iteration, delayed effects, and nonlinearities.

What I realized is that each of our group members' interests in a specific aspect of change represents a different dynamic associated with change. We're looking at cultural change from different perspectives. how we bring those together will depend on what we come up with in class next week. I do think, though, that defining a given system involves looking both at the necessary actors and the nature of the dynamics embedded within a system.

October 17, 2007

Fractal Music

Phil ThompsonStrange Attractors


I've been looking up different types of fractal music, and if you just go to www.youtube.com, and search for fractals, it will bring up various videos of Mandlebrot fractal, etc., some shown with fractal music accompaniment.
Also there is free software for creating your own fractal music online and sharing it at http://www.fractalmusiclab.com/.

In the book I've been reading "Complexification," it discusses what makes music interesting. "For any music to be interesting it should be neither too regular--like modern Rock, nor too unpredictable--like avant-garde compositions" (John Casti). In the book it discusses the different types of noise: white, brownian, pink or tan. White being purely random noise, while brownian has a correlation between the frequencies during successive time intervals. Pink or tan music is the edge of chaos and order allowing for the listener to no become overwhelmed or bored with the music. Music seems to be pushing forward with the switch from analog to digital but to me the analog music still has a warmth that digital hasn't been able to reproduce. Algorithmic music just doesn't have that emotion human touch that draws me in personally but that is a matter of taste I suppose.

October 8, 2007

Complexity, models , and learning

Our discussions of computer models of complexity and their relationship to the natural world are interesting, but there is another application of these types of models that does not engage the philosophical debate about what is real and what can be measured.

Computer models of complexity are used to help students (of all ages) think about how complexity might play out in solving problems that are systemic in nature. Researchers at the MIT Media Lab and their Epistemology and Learning Group have been trying to provide some tools that allow learners to play with these concepts. One product is a massively parallel piece of modeling software called StarLogo TNG. The support site (with some discussion of complexity) and free download are located at: http://education.mit.edu/starlogo-tng/complexity/complexity_and_SLTNG1.html

The site has some videos of simulations and some discussion of how complexity relates to the simulations. Although these simulations are of natural relationships, the researchers work to help learners understand that these models are adaptable and that many different models will show the same result. The goal is to use the model to help participants "think in terms of complexity".

It is interesting that this Adventures in Modeling project at MIT was a joint project of MIT and the Santa Fe Institute and that Eric Klopfer of MIT is a part of both institutions. A brief newspaper article about the project in Santa Fe Schools is at: http://www.plexusinstitute.com/services/E-Library/show.cfm?id=287

October 3, 2007

Discussion Addendum

Relative to the class discussion last Wednesday concerning the virtue of computer simulation, I would like to offer the following clarification of my thoughts.

The computer and its ability to simulate natural phenomena are significant additions to the tools with which thoughtful persons can search for the truth. Their ability to run multiple iterations of a process in a short time, together with their ability to change the input to each successive iteration based on results of prior iterations (i.e. feedback), and to do so according to fixed rules, facilitates the modeling of non-linear processes in a way that stimulates the mind and excites the imagination.

Such simulations demonstrate at least two propositions of Chaos and Complexity theory that I know of: 1) that complex behavior may be generated from the non-linear iteration of a few simple rules, and that 2) that a precise mathematical point may be found where an apparently linear process becomes non-linear (the “edge of chaos?). A 3rd proposition may be suggested by the 1st one: that all chaotic processes are only apparent, actually consisting of an orderly combination of simple elements.

These propositions are important and far-reaching in their implications, but they are rather abstract and general. In this respect, computer simulation differs from scientific experiment, which observes the behavior of real things. Binary strings and character strings are mathematical and/or logical constructs rather than natural phenomena, and so are the rules which drive the computer simulations. Thus, the “boids? simulation operates by rules of “average? flock density and “average? direction. Its results look tantalizingly close to how real flocks of birds behave, but they fail to prove that the rules which drive the simulation are what actually drive real flocks of birds. Conceivably, a different set of rules may achieve the same visual effect.

So, at this point in my 4-week acquaintance with Chaos and Complexity, I am inclined to think that the theory compliments traditional, reductionist science rather than encroaches upon its territory. No doubt, reductionist science has erred in offering itself as a comprehensive explanation of existence, and the holistic view of C&C is an important qualification of science; but until I see evidence that C&C theory is capable of results which supplant the results of traditional scientific experiment, I will tend to regard its propositions as more philosophical/mathematical than scientific. To the extent that its mathematical/logical constructs are truly representative of scientific fact, such as the rules governing DNA and chromosomes, computer simulation which observes those rules in action are a valid extension of scientific experiment; but without such a match, it is questionable whether computer simulation can proceed beyond the demonstration of C&C’s general propositions.

Mike Woolsey

October 2, 2007

Complexifying the Complexity Blog

What this blog needs is meta tags other than "administrative considerations", not that there's anything wrong with that.

Is there any way to make it possible to tag entries with things like 'nonlinear dynamics' or 'order out of chaos', or any number of the list of words David wrote on the board two weeks ago? Allowing these kinds of tags would go a long way towards making this blog less temporally linear, and more horizontally connected.

By the way, this particular entry clearly does fit under the category of 'administrative considerations'.

Posting to this blog...

Hello Everybody,

It has come to my attention that several of you have run into difficulty posting to this blog. I'll admit that I, too, find the user interface somewhat non-intuitive. Here's what to do:

go to http://blog.lib.umn.edu

once you enter your X.500 you will be taken to a list of all of the blogs you subscribe to. Click on LS 5100: Chaos & Complexity. This will take you to the MoveableType publishing platform. If you want to make a post find the "New Entry" tab on the left side of your screen. From there on its fairly self explanitory.

I Hope I'm not stating the obvious.