March 26, 2005

Deconstructing nonsystemic theories

Before deconstructing prevailing social interaction theories and theories about individual behavior, Hanson suggests an approach that bases research on starting with “wholes” or relational patterns. This is in contrast to traditional approaches of making truth claims about human reality that constrain the “range of phenomena” entered into analysis or used to frame a debate of issues. It means "seeing a world of relational wholes, rather than discrete individual pieces" (p.10)

Her basic idea is that conventional theories about social interaction or human behavior are premised on assumptions. The theory flows out of the assumptions. Hanson contends that rather than starting with a part (the assumption), it would be more helpful to start from the entire patchwork of relations between things. She points out that the atom bomb came from an epistemology that divorced science from ethics, yet the consequences of a thought (e.g. E=mc2) were fatal. She wants to use nonsummativity and multifinality and other "Wholes" concepts to make an epistemological shift from linear, compartmentalized short-term theorizing to a Wholes approach which includes cybernetics, relational unites and process. This doesn't discard old theories, it revises them. For example, conflict theory would be revised in the "language of equifinality to describe the inevitability of conflict" (p.10)

Social theories

Conflict theory
• Assumes human goodness & that economics drive society
• Based on Heglian dialectic, i.e. a struggle between opposite forces—the haves and the have-nots
• Based on view of macro systems, e.g. social structures
• Goal is to raise collective consciousness through awareness of historical analysis

Consensus theory
• Assumes human greed
• Social structures arise to contain “insatiable human desires”
• Parts of system help maintain homeostasis
• Competition dictates which person/skills get the more or less advanced parts
• Consensus about organization of the whole, although conflict occurs in areas of minor deviations within whole

Symbolic Interactionism
• Assumes humans are creative & able to reflect on their behavior
• Humans share meanings (symbols) about what is important, e.g. language
• Focus on micro level
• Subjective meanings of the self determine self-purpose, membership in groups, etc.

The micro/macro debate
• Debate between micro vs. macro OR interpersonal, subjective, personal vs. societal, objective, abstract.
• Assumes that either micro or macro has greater causal influence
• Assumes a linear causality and, therefore, blame
• Research--What is the unit of analysis? Raises the aggregate vs. context issue…this is maybe more salient than objectivity vs. subjectivity
o Aggregate—the societal or general social property is analyzed; data from individuals and generalized for whole
o Context—the particular and subjective are analyzed; data of majority response is contrasted with the largest proportion of a unified response.
Implications for building a research model:
• Aggregation to the universal may bias the most frequent response even though it’s not the majority response
• Even when the focus is on the majority, it may miss the subjective views of particular groups
The issue of [focusing on the particular context vs. proposing a generalization via aggregation] shows one way that “we move from the theoretical to the practical and what is means in terms of interpreting results.”
• Does the average represent the particular?
• Does the perceived statistical predominance represent consensus, or a lowest common denominator, or the views of the most powerful?

Theories of Individual Behavior:

• Humans and animals share similar characteristics
• Humans can be reduced to utilization of artifacts through series of behavioral rewards/punishments
• Based on logical positivism & nomotheism (deriving universal properties based on the observation of individuals & then aggregating these data into average properties).
o So, you can generalize from individual to general & from animal to human

• Assumes human subconsciousness (id, ego, superego)
• Goal is to dig into unconsciousness and release pent up frustrations
• Based on logical positivism—focuses on objectivity of the analyst although the response is to the subjective meanings of the human subject

These theories require you to accept their underlying assumptions. Because each approach is belief-based (beliefs form the foundation of the theory), it cannot be reduced beyond the basic assumptions. Any contrasting assumption must be rejected. This is how the classic debates proceed. However, starting with the whole first has promise to formulate a different process.

General systems is really an approach, not a theory. The beginning point of the approach is not the basic assumption, it is the “point of departure, nonsummativity, which states that the whole is greater than the sum of its parts.” When you start with a Whole, it is not just an adding up of all the individual variables, it is literally a different animal.
There are two ways to view the current war in Iraq, as an illustration. One is that the war had a beginning and an end..."mission accomplished." This linear, short-term view only needs to look at a beginning and end point in order to determine whether the goals/outcomes were achieved. The war is a containable thing. In contrast, a view of nonsummativity would gather all variables and look at them as a Whole. From the preparations for war (impact on American economy/conscience) to the dovetailing effects of a regime change in Iraq upon the Middle East, Europe and the rest of the whole (precedents, ideologies, societal trends, etc.). Hanson puts it nicely, "a wholes approach means seeing not just the initial effects but how these effects are reacted to, how the process amplifies and mutates from the original" (p. 12).

So, here's my question. If we start with the Whole, do we end up with a thousand parts/pieces?


Yes...maybe more.

If you begin with the BIG PICTURE, you can make sense of an infinite number of small details. To put it epistemologically, the systemic/ecological approach necessarily requires model-building; it is not possible to communicate in "pure" systems terms, other than mathematically (which is also modeling, but at a very general level). All of the theories you have summarized above are models examining certain aspects of reality (and of necessity neglecting others, even though we know they are "there"). Modeling permits selecting a certain number of parts/pieces from an infinite number and focusing on them for the purpose of "simplifying" for understanding. As Weinberg puts it, models allow us to talk about something we are trying to understand in terms that we think we already do understand.

Posted by gschache at 10:38 AM

March 8, 2005

(Probably) Simple Questions

I feel like an infant joining this discussion -- I'm not working on the same level as you, Gregg, and it's going to take me a while to catch up. But I haven't been idle: I've been playing with these ideas seriously enough to have (for one brief and fleeting week) announced to Bill Doherty that I was changing the direction of my graduate work and that I'd do a theoretical dissertation on an idea which systemically addressed an intersection between quantum physics, interrelationality, and living systems. I had even started assembling a committee around it (Liz was very interested), when friends from other institutions cornered me at a conference and told me I was talking about life's work, not a dissertation. Or at least not one which could be completed in my life time. Maybe yours, Gregg, but I'm SOOO much older than you.

Bill was very relieved when I changed my mind.

I'll begin to start dialogueing with you both about this paradigm shift I'm playing with, but not today. I thought I'd start with really basic issues, like how systems connect.

It seems from my reading that systems are often presented in a vertical or hierarchical manner. If I could be sure this would tranfer correctly, I'd try to diagram it for you, but I think I'm stuck with a written description, so here goes:

Take a single individual. Call him System 0. Move upward, or outward (if you'd like to think of shells rather than stay on a linear plane): above System 0 are increasingly larger systems (say, System 1, System 2, System 3, etc.). These macrosystems have a commonality in that each has the individual (System 0) as a member. For example, System 1 may be a dyad with his wife, System 2 his nuclear family, System 3 extended family, System N community, or society, and so on. There may be macrosystems which do not include all members of other macrosystems in which System 0 is a member, such as one macrosystem being his church, another his workplace; hence, a shell scheme. But still, they are all of increasing size and complexity, and all have System 0 as a member.

Now, we'll work in the inverse direction. Back to System 0: He is comprised of microsystems, each of smaller composition, each a member of System 0. So, we've got the respiratory system (System -1), bronchioli (System -2), lung cell (System -3), cell organelles (System -4), proteins (System -5), and so on. Again, there is decreasing size and complexity, but all have System 0 as a level of the hierarchy.

Am I seeing this correctly? Making sense?

Now, my question is if Systems Theory addresses lateral relationships, one in which there may be a connection to System 0 at some distant macro level, but there is none in linear proximity at the point of connection.; rather, one can view these more as lateral or horizontal connections. Let me explain more, as I struggle to language this. Take System 0 and a germinating grass seed (another System 0). Research has been done to indicate that a human can project a sense of wellbeing onto a germinating seed and there is an observable increase in the rate of plant growth. How can you explain this is System Theory terms? Would one say that the human is part of the seed's ecosystem (on a macrosystem level) and that is where the interrelation takes place? Or can independent, vertically unconnected systems impact each other? (Duh. Of course they can, as I reread this. But how to describe it correctly within the theory?)

As I write this, I'm beginning to see it a little more clearly, but I'd surely appreciate some input. Thanks.


Mary, you are grappling with the deepest levels of the systemic paradigm. First, I concur with your friends that defining ultimate reality is a bit ambitious for a doctoral dissertation. I was warned in graduate school (and I needed the warning) that I probably should not try to write my magnum opus in a dissertation; it was a struggle, but I managed to become more cirumspect over time. Nevertheless, the issues you raise can--and will--inform your graduate work throughout, as well as future professional endeavors. There are two REALLY BIG issues reflected in what you have written. First, the ontological one: Are systems "arranged" hierarchically--and, by extension, what is the "real" nature of relations between systems in the universe? Second, the epistemological one: How can we know (understand), language, and depict the relationships between systems? I studied with Paul Tillich at the University of Chicago, and he had a very tightly argued dialectical ontology that talked both about "dimensions" of life and "levels" of reality. I would be glad to loan you a couple things that you might find interesting. The epistemological issues have been addressed primarily by Bateson, and, more recently, by Humberto Maturana. I also have a prepublication article I have been working on for a decade on "Depicting Ecosystems." I have presented it at a couple of conferences and to classes here in the Department. But it has never seemed "complete" and ready for publication (I think I will always feel that way, so I have decided to give it one more edit and then submit it; if no one understands or likes it, I can say it reflects my senility now that I'm retired). One other simplistic comment: Systemic relationship seems to be most commonly presented as some sort of concentric circles (you might remember the sexual development diagrams by Robert Francoeur that I handed out in the couples course). I do believe that such depictions are more useful that a diagram of "levels," although they have some limitations, too; ie., that systems on the outer rings are "larger" than those on the inner rings.

Posted by mkellehe at 3:41 PM

March 5, 2005

Insights from Cybernetics

Chapter: "On the relation between cybernetics and general systems theory"
Author: George J. Klir

There are a couple features of cybernetics that seem to have widespread applicability across disciplines. One aspect is that cybernetics deals with communication and control. These two areas could apply to humans, their societies or even the arts. Another aspect of cybernetics is information. This may be the area most closely commonly associated with cybernetics. Without information, communication and control are shipwrecked at the onset.

In cybernetics, information is something that can be quantified and measured. Klir defines information as, "a measure of organization as opposed to randomness" (p. 157). As such, the amount of information being inputed/outputed depends the QUANTITY, as well as, QUALITY of organization. Klir goes on to point out the dangers of ignoring either the quantity or quality of an organization. Neglecting quality leads to "negative entropy" between information and organization. Klir does not point out a danger of neglecting quantity, but I would imagine it has to do with limited impact, i.e. super simple systems of information with little opportunity to evolve. Am I offbase with this intuition?

Klir goes on to show how cybernetics can serve as a viewpoint or as a discipline in understanding ifferent disciplines. A way that cybernetics would be used as a viewpoint would be in understanding the process of metabolism as an interaction between mass and energy. So, examples of what might be considered cybernetically would be the amount of information (quantity) examined or the evolution of information within the system/between system and environment.

So, what are some application points that might be envisioned? In building a model that would examine a vast social system, I would think that one danger might be sacrificing quality data (using census data to the exclusion of intimate interviews of families or groups). The other danger might be a model that only examines phenomena on dyadic or triadic levels within a family, rather than considering the entire ecosystem. I suppose this dual emphasis in model-building could be credited as a cybernetic principle. Would this be an accurate way to think of it? BASICALLY, YES.

Another application point comes from Klir's example of examining the evolution of information between a system and its environment. This implies that there are things larger than or distinguishable from systems that systems can interact with. So, another way of classifying systems and their environment would be needed to comprehend this interaction. Is that where the subsystem vs. Ecosystem distinction derives? YES!


Your reading of Klir, and Klir's work itself, nicely portrays the dialectic tensions between epistemology and ontology, as well as (in systems terms) between structure and process. Note that Klir refers to cybernetics (which is by definition process-based) and then discusses issues of organization of a system and the relationships between systems and their environments, etc. as if they were "real THINGS." You can see what I am convinced that dialectics are centrally embedded in systemic thought--without such a philosophy and heuristic tool, it would be difficult to make sense of the relationship between cybernetics as information processing and general systems theory as a theory and study of living and nonliving systems.

Posted by gschache at 4:59 PM