Talk:Complex socio-ecological systems/Complex Adaptive Systems

SUMMARY DISCUSSION We did not concentrate on a particular paper but discussed concepts that were common to both. -Before diving into addressing the questions posted, we dealt with the issue of linearity vs. non-linearity. The concept of threshold was discussed, and refers to the case when the simple linear relationship between two variables is not sufficient to explain the novel configuration that a system can take. -Emphasis was made on the importance of feedback both within the CAS and between it and the surrounding environment. -The importance of history of the CAS (path dependency) was made. The evolutionary process has often been explained through the analogy of a random walk. Once a path is followed, the possibilities are different from those from a path taken on a different direction. Advancing through these “steps” determines (constrains) the possible configuration of CAS, but also provides novel opportunities for innovation and change (learning, adaptation). - There was some discussion about how could culture and tradition be considered, as if providing more or less resilience of social systems to change. We talked about mental models in the context of indigenous knowledge, and how adaptation could not be necessarily interpreted in a positive way (good vs. bad adaptation). - We discussed what were the differences between complex systems and complex adaptive systems, and some examples of the latter came out (global economic system). - We talked about basins of attraction and how the shape of these are determined by the diversity (variance) of potential configurations a system can take, and how they relate to resilience. The analogy was made with income sources for rural communities: the more they depend on a single resource (narrow basin of attraction) the more vulnerable they are to unexpected changes (surprises) of the CAS and the environment. -WE made emphasis on the existence of a transition zone between extreme systems’ configuration. This was discussed with the example of longleaf pine savannas and hammock forests in Florida. These two extremes represent stages where fire is following its natural regime (low intensity frequent fires) in long leaf pine savannas, to places where this disturbance has been removed. Yet, in nature we find a series of configurations along that gradient so it is important to remain flexible when trying to understand CAS. -The previous discussion was important when addressing the issue of a system being “at the edge of chaos”. Given that Levin and Lansing recognize the value of being in this stage as it provides the best opportunity for innovation and adaptation, it was important to try to bring back the issue that systems are not or frozen or chaotic, and there is a space between periodicity/stabiity and chaos that is important to consider. -It is important to keep the issue of scale in mind. Although we did not have enough time to address the five questions posted, it was interesting in the sense that we are all struggling to become more comfortable with the CAS ideas.

Claudia 14:17, 27 January 2011 (UTC)

Hi! WE thought we would send some questions/issues/thoughts to think (even) further… to spark/guide some discussion tomorrow.

Issue 1. “Heavily managed systems, such as in agriculture or forestry, are not purely complex adaptive systems, in that their simplified structures are imposed exogenously rather than arising endogenously” (From Levin, 1999).

Anthropogenic is always exogenous? Why is this important when looking at CAS?

Issue 2. From Lansing paper there is the suggestion that closeness to the edge of chaos means more adaptation potential. What are the costs of being at the edge of chaos? Implications of this closeness?

Issue 3. Comment this statement…P. 189 (Lansing). “At the other extreme, when K is small, networks exhibit stable or periodic behavior”.

Issue 4. “…To what extent can the patterns of distribution of ecosystem properties be explained by under lying variation in physical variables, such as regional climate and soil conditions, and to what extent are they the result of self-organization”. (From Levin 1999). Having been exposed to all the talk about systems from last 2 weeks what would you say is missing from this statement?

Issue 5. “.. Aggregation and hierarchical assembly are not imposed on complex adaptive systems, but emerge from local interactions through endogenous pattern formation..” (From Levin 1999).. Any comments on this?

claudiaClaudia 19:50, 20 January 2011 (UTC)

comments
Hi! Please remember to post comments today. One of us has not internet access after 5 pm.. Thanks! C&C Claudia 14:35, 20 January 2011 (UTC) - Lansing and Levin’s pieces compliment each other in surprising ways, and not merely because they both discuss CAS in evolutionary biology, but because they speak to the existence of emergent properties. This is the fundamental agreement I seem to find in all this literature, that scale is of the utmost importance in deciding when a system is linear or not (and most of the times it’s not). Amid all this discussion of basins, K states, and epistatic convergence, Lansing’s critique of CAS is more robust than Levin’s in that he takes a grand survey of the different fields where CAS has been applied. But it is Levin who is clearer in his delimiting of where CAS and sources of resilience (or the edge of chaos) can be found. Lansing infers that there is a separation between adaptive systems like cellular automata and complex, but arguably nonadaptive systems, like post-disaster landscapes. I guess this illustrates how dated or unenlightened my reading of complexity science has been, but I somehow assumed that complex systems and CAS were one in the same, that both took into account nonlinear processes, and that the social scientists had led much of the way alongside ecologists in nonequilibrium thinking. This is apparently not the case, at least according to Lansing and many of his peers, but one can’t help but pick recent editions of Ecology and Society and other journals that very much do speak of adaptation from both acute and chronic shocks and stressors, in ways that do not illustrate the kind of self-organization Lansing endorses. This literature hails itself as the work of adaptive science. Admittedly I’m waist deep in it so perhaps I need to step back and take a more critical eye in the same way Lansing might. The dissonance in his paper is a little bit troubling for me, though, just as the postulation that artificial intelligence has been biased towards the science fiction, when its real world applications (like the prisoner’s dilemma, or modeling of Bali rice fields) should better steer complexity science—aren’t both a reflection of Levin’s “anthropic principle” in some way (this goes back to last week’s discussion of models and how all of them are “wrong,” or at least biased in some form). I think this would be a great jumping off point for discussion on Friday if the format hasn’t already been decided upon. 68.105.165.243 15:59, 20 January 2011 (UTC) Sam

Lansing:

It was nice to read about how debate and research on complex adaptive systems evolved and the different concepts associated with this field of research. Some concepts were new to me, such as "basins" of attraction and "edge of chaos".

One element that caught my attention is the discussion regarding the fitness of an individual versus the the fitness of a network of connected individuals. What I understood here is that the more connected an individual is, the more it/he/she will depend on the other individuals to evolve: "the fitness of an organism or species depends upon the others which with it interacts" (p.186). However, Lansing also explained that the number of connections (K) affects the adaptive landscape, changing it from "Mount Fitness" to "tiny hills." Does this mean that the more organism are connected, the less adaptation is needed? This was a bit confusing to me. I get how more connections make it more difficult for any individual to evolve by itself, but I don't understand how connections affect the adaptive landscape.

The impact that the number of connections have on the behavior of the system, and its capacity to adapt: Based on the author's explanation of the Kauffman/Langton experiments, I understood that there is an ideal number of connections that allows information to be transmitted and the system to adapt. Few connections can result in a frozen system and too many connections can result in a chaotic system. The ideal state when thinking about adaptation is to have around 2 connections, which would make the system be at "the edge of chaos," a state that allows for adaptation and information transmission.

I find it somewhat difficult to grasp that complex systems are at the "edge of chaos," which is a good state because it allows for adaptation. The implication of this idea is that societies need to be at this state for it allows societies to adapt and get to "food sources of higher quality." Not being at the edge of chaos would mean that a society is in either a chaotic or a frozen/periodic state, which are worse states because they do not allow information flow and adaptation. The implication is that complexity (the edge of chaos) is good for societies, because it allows evolution.

Basins of attraction: How is the "basin of attraction" different than an equilibrium state? The idea that lights blink disorderly, but eventually go to the "state cycle or pattern of twinkling" tells me that systems tend to move to an equilibrium state.

The discussion about ENSO/Borneo forest talks about self-organization of systems and reminds me of ecosystem "equilibriums." However, there is a move away from the idea of equilibrium in nature, as exemplified by the author's "critique of equilibrium theory."

I find two ideas to be contradictory: 1) On one hand, repeated interactions between humans can lead to optimal outcomes, such as the Balinese irrigation systems. 2) On the other hand, humans can really mess up natural occurring self-organizing systems, like the policy makers in Borneo, who did not take into account the self-organization of the forest. Why interactions between humans and between humans and ecosystems led to a positive outcome in one case but not the other? Is it because in Borneo there were NOT repeated interactions (which are important for learning and collaboration)? Will the better outcome eventually come in Borneo, since now policymakers there know the negative feedback of their actions?

In blaming environmental problems on the excessive number of people in the world (as the author does when he agrees with Holland), the author seems to assume that humans are not part of self-organizing systems. The message seems to be: Humans, an exceptional species, mess up self-organizing systems. Doesn't the self-organizing principle applies to humans and human society?

Emergence of institutions: Author points out that institutions can emerge from the bottom-up as a result of local interactions. The image I got was of institutions popping-up unintendenly as a result of local interactions -- in an unplanned way. I believe the author did not meant to portray this imagine, but only that institutions can emerge from the bottom up as social actors learn from the feedbacks they receive. I guess the word "emergence" is the word that makes me think of "unintened" institutions.

While this is the case sometimes (bottom-up institutions), as the literature on the commons show, most  often institutions result from planned action to serve specific functions, such organizing people/processes in order to achieve specific outcomes, or as a means of social control. And institutions are often created top down, rather than allowed to "emerge". The Balinese case brings up the question of whether it would not be better for policy makers to allow institutions to emerge as a result of feedbacks and local action? This would mean allowing the learning and cooperation to occur. I know... I always bring these "should" questions. I guess I can't avoid thinking about policymaking. After all, aren't we learning about complex systems because we assume that by understanding systems and complexity we can make better decisions? flaleite 21:46, 20 January 2011 (UTC)

Sorry this is late, everybody- I spent the day in bed with food poisoning. Disclaimer: Hello, everyone! I suppose I'm posting as today's resident anthropologist, but you should by no means take me as a representative sample as a)I'm a graduate student, and b)I'm a product of my training, which is from an applied medical anthropology/biocultural/fairly materialist point of view. So my take on the readings, and thus my questions, come from that standpoint, and from an ongoing assessment of what is, and is not, useful to my own work.

Field-Specific Background: Just for some quick "History of Anthropology" background on the readings, the Whiting six-culture study mentioned in the Levin paper was a landmark paper in anthropology, and a comparative paper that created a whole new generation of scholars. John and Bea Whiting were students of John Dollard, who, along with Edvard Sapir, George Peter Murdock (from the Murdock "Standard Cross-Cultural Sample" also mentioned in the Levin paper)and Clark Hall, were part of the "Yale Synthesis", which, to my best guess, was simply a group of graduate students who got together and rejected the uber-relativist configurational school of culture and personality spearheaded by Margaret Mead, Ruth Benedict, and Franz Boas. To the best of my knowledge (and remember, I'm still a student) the Yale Synthesis was a theoretical perspective that rejected the idea that cultures could not be compared along material lines. One of the many brainchildren that came from this school was the HRAF, or Human Relations Area Files, created by George Peter Murdock (who went on to found the Department of Anthropology at the University of Pittsburgh), which was one of the first attempts made to quantify ethnographic data for purposes of comparison. The HRAF files have been critiqued heavily for many of the same issues that Levin mentioned as being problems with his dataset (Levin, pg.5-6); regional biases within the dataset, unequal training in/focus on demography among ethnographers, problems with sampling (including culture unit size and missing data), and a lack of systematized collection. Many of these problems, as far as I can see, are at least initially inescapable given the way that ethnographic work is currently executed. Many different ethnographers = many different goals and many different foci. It may be difficult or inadvisable to use an ethnographer's demographic data on child rearing practices when the ethnographer in question spent their field time focused on religion, and gave child-rearing and household ecology only the briefest of nods in their dissertation.

Questions for Myself and Others: However. Stating that there is difficulty in doing a thing is a far cry from saying that it cannot be done, and I feel there is danger, at least for applied fields, in assuming a point of view so relativistic as to not allow for any comparison between cultures whatsoever. This brings up what seems to be one of the core questions of our ongoing discussion; assuming complex systems can be modeled, can be compared in meaningful and useful ways, HOW is it to be done? Can we take steps- or better yet, should we take steps- in the very infancy of our research to make it more systematized and more palatable to comparative studies later on?

Also (and I'm trying to keep things basic, as class discussion so far had been fairly theoretical);

Can there be a happy marriage between ethnography and demography? How can the difference in traditional scales and modus operandi between the two fields be reconciled? Take, for example, the numerous methodological problems encountered in the Levin, 1991 dataset.

Given the differing goals of different ethnographic studies, is it good practice to use demographic data from multiple studies who may or may not have been interested in the same topics, and therefore lack any form of unified systematic data collection? (Think about the Pitt/Yale Human Relations Area Files (HRAF)). How can we get around this? Cbaird