The Systems Thinking Wars

My goal for The Multidisciplinarian is to talk about multidisciplinary and interdisciplinary problem solving. This inevitably leads to systems, since problems requiring more than one perspective or approach tend to involve systems, whether biological, social, logical, mechanical or political.

I hope to touch upon a bunch of systems concepts at some point, including:

I started following some of these terms on Twitter a few weeks ago, and ended up reading a lot of web topics on Systems Thinking. I found all the classics, along with, surprisingly, something of a battleground. I don’t mean attacks from the outside, like the view that organizations are not systems but processes. Instead I’m talking about the enemy within. It seems there are several issues of contention.

The matter of whether Systems Thinking is a deterministic or “hard” approach percolates through many of the discussions. “Hard” in this context means that it’s a mere extension of systems engineering, treating humans, society, and business organizations as predictable machinery. But on the street (as opposed to in academics), there’s also disagreement over whether that attribute is desirable or not. Some proponents defend Systems Thinking as being largely deterministic against criticism that it is soft. Other defenders of the approach argue against criticism that it is deterministic.

Is Systems Thinking an approach, a model, a methodology, or a theory? That’s debatable too; and therefore, it’s being debated. One can infer from the debates and discussions that much of the problem stems from semantics. The term means different things to different communities. Such overloaded terminology works fine as long as the communities don’t overlap. But they do overlap, since systems tend to involve multiple disciplines.

From a distance, you can grasp the gist of Systems Thinking. At its most rudimentary level, it is seeing the forest from the trees and using that vision to get things done. Barry Richmond, celebrated systems scientist, gave this high level definition:

At the conceptual end of the spectrum is adoption of a systems perspective or viewpoint. You are adopting a systems viewpoint when you are standing back far enough—in both space and time—to be able to see the underlying web of ongoing, reciprocal relationships which are cycling to produce the patterns of behavior that a system is exhibiting.

Peter Senge of MIT says that Systems Thinking is an approach for getting beyond cause and effect to the patterns of behavior that surface the cause and effect, and further, for identifying the underlying structure responsible for the patterns of behavior. If you, perhaps recalling your philosophy studies, detect a degree of rejection of reductionism in that definition, you’re right on track. More on that below. See the Systems Thinking World‘s definition page for a list of other definitions.

Barry Richmond, like Jay W Forrester, his mentor and prolific writer on Systems Thinking, was also heavily involved in System Dynamics. While many people equate the two concepts, others distinguish System Dynamics from Systems Thinking by the former’s use of feedback-loop computer models. Forrester, a consummate engineer and true innovator, developed the Systems Dynamics approach at MIT in the 1960s.

Caffeine Systems Thinking
System dynamics model showing processing of caffeine by the body and effects on drowsiness

For several decades Forrester applied Systems Thinking to business management, society and politics, maintaining throughout, that system dynamics is the necessary foundation underlying effective thinking about systems. In a 2010 paper, Forrester, then in the Sloan School of Management, wrote:

Without a foundation of systems principles, simulation, and an experimental approach, systems thinking runs the risk of being superficial, ineffective, and prone to arriving at  counterproductive conclusions. Those seeking an easy way to design better social systems will be as disappointed as if they were to seek an effortless route to designing bridges or doing heart transplants.

These bold and beautiful words are lost on the those who only know systems thinking from its current usage as little more than a strategic-initiative group-hug word. The quote is from Forrester’s appeal that Systems Thinking, at least as popularly defined, is insufficient without system dynamics modeling. Forrester speaks to usage of Systems Thinking that is nearly as deflated as current usage of “six-sigma,” by which our ancestors meant standard deviations of manufacturing tolerance (statistical process control). Nevertheless, as sociolinguists point out, a word means what a large body of its users think it means.

In the spirit of multidisciplinarity, it’s tempting to view this war from the perspective of study of religious cults. Too tempting – so I’ll succumb.

As with the internecine battles of religious cults, this is a war of small differences; often the factions in greatest dispute are the ones with the most similar views. Their differences are  real, but  imperceptible to most outsiders. They argue over definitions and interpretations, engaging in doctrinal disputes with constant deference to the cults’ founders. I also detect a fair amount of anxiety of influence in Systems Thinking advocates with roots in hard sciences.

Many systems engineers, including some very good ones, after opening the door to systems thinking, strain to differentiate themselves from their less evolved brethren.  John Boardman and Brain Sauser, thought leaders for whom I have the utmost respect, oddly display the anxiety of influence in statements like this from their Worlds of Systems site:

Our engineering friends believe the term ‘system’ is theirs of right and they alone understand systems. After all, who builds them? Who gets the job done? You would think, to hear some engineers talk, that they invented the term itself. In fact what propelled it into the high currency values it occupies today were the ideas of Ludwig von Bertalanffy.

Here we have two brilliant engineers (see in particular their work on Systems of Systems) who – though perhaps in jest – downplay the development of systems thinking a la Forrester, deferring to Bertalanffy, the biologist who first used the term Systems Theory. Semantic mapping tools available on the web clearly show that Bertalanffy, ground-breaking as he was, had next to nothing to do with the propulsion of the term “system” to its current status. The route was, as you’d expect, from Greek philosophy to Renaissance astronomy, to biology and engineering, and then on to computers.

Without delving into heady problems of Bertalanffy’s worldview, such as the paradox of emergence and the paradox of system environment, I’ll suggest that Bertalanffy was a great thinker, but should not occupy too high a pedestal. His view that the reductionist nature of biology of the mid 1900s stemmed solely from the influence of Descartes and Newton (who thought nature could be modeled as mechanism) ignored the obvious necessity of reduction in order to link stimulus with response. Testing ten foods separately, to see which causes your allergic reaction, does not conflict with holism. Bertalanffy, despite his great contributions, beat a reductionist straw man to death. Finally, can anyone not find Bertalanffy’s language of his later works indistinguishable from that of liberal theologians? Paul Tillich meets business management?

Boardman and Sauser similarly quote Philip Spor’s remark, “the engineer must often go beyond the limits of science, or question judgment based on alleged existing science,” as if such going-beyond isn’t inherent in engineering. Really guys, does anyone really think that the science of turbomachinery predated the engineering of turbomachines? Recall that special relativity was solid before the fourth-order partial differential equations governing a turbocharger were nailed down, at which time Alfred Büchi ‘s invention was common on trucks and trains. The opponent here is also mostly made of straw – a purely reductionist caricature of a systems engineer.

As a scholar of history of science and a fan of history of religion, here’s what I think is going on. Systems thinking is often at the intersection of systems science and social and management science; and the most orthodox of each of those root beliefs accuses the others of being too hard (as seen by social science) or too soft (as seen by engineers). The most liberal (or reformist, in the religious model) accuse their own party of being entrenched in orthodoxy.

Cult members mine the writings of these clergymen for ammunition against rival cults, thus we see quotes from Forrester, Bertalanffy, Ackoff and the like on websites, grossly misunderstood, and out of context. And we see ludicrous and undisciplined extensions of their material, as with Gary Zukav, Fritjof Capra, and Roger Penrose. The cult’s most vocal advocates insist on deifying the movement’s founders, and speak in terms of discovery and illumination rather than evidence and development.

Reasoning by analogy, yes; but I think you’ll admit this analogy holds rather well.

Another face of the Systems Thinking wars deals not with definitions and philosophy but with efficacy. In a 2009 Fast Company piece Fred Collopy, an experienced practitioner and teacher of Systems Thinking opined more or less that Systems Thinking is a failure – not because it has internal flaws but because it is hard. Systems Thinking, says Collopy, requires mastery of a large number of techniques, none of which is particularly useful by itself. This requirement is at odds with the way people learn, except in strict academic circles. Collopy offers that Design Thinking is an alternative, but only if we can keep it from being bogged down in detailed process definition and becoming an overly restrictive framework. He notes that if Systems Thinking had worked like its early advocates hoped it would, there would be no management-by-design movement or calls for integrated management practice.

Interesting stuff indeed. It will be fun to see how this plays out. If history is a guide, and as Collopy seems to suggest, it may fizzle out before it plays out. Business schools and corporate leadership have a record of moving on to new, more fashionable approaches, independent of the value of current ones. More on that tomorrow.

———————–

Philosophy of science is as useful to scientists as ornithology is to birds. – Richard Feynman

Thanks to Ventana Systems, Inc. for use of their VENSIM® tools.
Thanks to @DanMezick for recent tweet exchange on Systems Thinking.

, , , , , , , , ,

  1. #1 by tomfid on March 26, 2012 - 10:52 pm

    I’d say this is a pretty accurate characterization of the debate that I’ve seen simmering for 20+ years. Perhaps not surprisingly, it’s a lot more obvious when you’re in an academic crowd, or among practitioners without their clients. On a consulting engagement with a real organization, typically nobody cares much about philosophy. Personally, I’m on the hard side, but I think the two camps need each other – ST needs modeling for a reality check on errant intuition, and modeling is nothing unless people will implement it.

    I think the “ST is hard” argument has a lot going for it. As long as organizational and market incentives are short term, the payoffs to SD/ST are remote in time and space, and it’s hard to make causal attributions of success and failure, there’s evolutionary pressure in favor of unsystemic behavior. That is a recipe for organizational and planetary disaster of course.

  2. #2 by turnockopinion on March 30, 2012 - 8:38 pm

    The systems concepts can also be divided into qualitative and quantitative. Examples of qualitative tools include causal loop diagrams, connection circles, lists, cause and effect diagrams, and many other verbal-focused qualitative tools and concepts. Extroverts and verbal communication personalities focus on qualitative tools and concepts. They tend to dominate the literature, presentations and consulting engagements. Mistakes are easily made using qualitative tools and concepts. I label qualitative systems tools as Systems Thinking.

    Quantitative systems tools include dynamic modeling using Stella, iThink or Vensim. The concepts of stocks and flows, feedback loops, cause and effect, making assumptions explicit, understanding how structure determines why a system behaves the way it does and other concepts require quantitative dynamic modeling skills. I label quantitative systems tools as System Dynamics.

    I saw your post on Linkedin the systems thinking world discussion.
    RKT

  3. #3 by Gene Bellinger on March 31, 2012 - 6:48 am

    Bill,

    Very nice piece. Thanks for connecting it to me. Sorry there were so many interesting links in it as I now have several other things to read I don’t know where I’ll find the time for. Catch you in STW.

    be well,
    Gene Bellinger
    Host Systems Thinking World

  4. #4 by Adrian Chan (@gravity7) on June 7, 2012 - 10:54 am

    Nice summary. I would add Luhmann’s social systems perspective, with its emphasis on auto poesis and communication in particular. Systems reproduction, when it’s not physically causal but relies instead on people, is caught up in matters of meaning — in action and communication. Philosophical issues of interpretation by actors of events and of actions become unavoidable.

  1. Management Initiatives and the Succession of Divine Generations « The Multidisciplinarian
  2. Science, Holism and Easter « The Multidisciplinarian

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: