Qualitative/Narrative Complexity Science

For all of its advances (and they are many) complexity science has yet to bridge fully the rift between qualitative and quantitative method.

Before I explain myself, however, some quick definitions are in order. First, by qualitative method, I mean the non-numerical analysis of narrative and verbal data, as typically studied in historical inquiry, ethnography, qualitative interviews, and grounded theory. By quantitative method, I mean the study of numerical data, primarily through the application of statistics and top-down equation-based modeling.)

To its credit, complexity science has significantly progressed the qualitative analysis of numerical data. By "qualitative analysis" I mean the study of the complex, emergent, relational, dynamic, evolving, idiographic dimensions of numerical data. In fact, one could claim that complexity science method is really a major advance in the qualitative study of complex numerical data.

What complexity science has not advanced, however, is the non-numerical study complexity. To date, only a handful of articles have applied qualitative method to the study of complexity. And even fewer articles have examined how to advance the usage of qualitative method for studying complex systems.

The earliest examples I know of that apply qualitative method to the study of complexity were written by Crabtree and colleagues (most of whom are in medicine, nursing or health finance) and their study of medical practices:

1. Crabtree, B. F. (1997). Individual attitudes are no match for complex systems. Journal of Family Practice, 44(5), 447-448.

2. Crabtree, B. F. (2003). Primary care practices are full of surprises! Health Care Management Review, 28(3), 279-283.

3. Crabtree, B. F., Miller,W. L., Aita,V. A., Flocke, S. A.,&Stange, K. C. (1998). Primary care practice organization and preventive services delivery: Aqualitative analysis. Journal of Family Medicine, 46(5), 403-409.

4. Crabtree, B. F., Miller,W. L.,&Stange, K. C. (2001). Understanding practice from the ground up. Journal of Family Practice, 50(10), 881-887.

The earliest (and most widely popular) example of the development of qualitative method for the study of complex systems is Charles Ragin's Fuzzy Set Social Science (2000). Ragin also has a new book with David Byrne (a prominent British sociologist and leading scholar in the social science application of complexity science--I will blog more about this book later). The title of the book is The SAGE Handbook of Case-Based Methods (2009).

Despite being a small literature within complexity science, these scholars make some very compelling arguments for developing the qualitative (non-numerical) study of complexity. Perhaps the best argument is that a significant amount of data goes unexplored when qualitative method is not used.

What, for example, are the phenomenological dimensions of complex networks? What does it mean for people to be connected to one another by six or fewer links? What are the emotional dimensions of being part of a massive online social network? What role do power, conflict, hate, greed, anger, and love play in the complex global system? How does one study "confidence" in a system? What does a state of domination within a complex social system look like? Is altruism within a system more than a prisoner dilemna? I could go on and on and on.

Okay, just one more example: Think about the current global financial collapse in which most (if not all) the world is struggling? How do people make meaning of this experience? And, to consider second-order cybernetics and sociocybernetics, what consquence does the meaning people make have for the way in which our global economic system will evolve? And so on and so forth.

There is a lot qualitative method can offer complexity science. And, there is a lot complexity science can offer qualitative method. If complexity scientists turned their attention to this dimension of method, they could create some very incredible tools.


Rockin' Mandelbrot Song


This is absolutely the coolest math song ever written. I gave a MATH DAY presentation about two weeks ago for 300 math geeks and they went crazy! It is fanstastic. Play it for yourself, friends, profs, and students--especially students in the social sciences and humanities.

The song is by Jonathan Coulton. The video was made by Pisut Wisessing in Film 324: Cornell Summer Animation Workshop, taught by animator Lynn Tomlinson.


Dungeons & Dragons--the Geek Stereotype

Okay, so most of us geeks fit the stereotype--instead of going on dates in highschool with humans, we were dating elves (male or female) or any other assorted group of medieval characters. D&D anyone?

We geeks eventually grew out of this phase. Actually, no we didn't--which brings me to the point of this post. One of my geek buddies (Michael Ball) has gone and done the worst thing a medieval geek can do. He wrote a book about it.

Mike's first fiction book is titled The Stone Men. It is an excellent short story with fantastic illustrations drawn by Christopher Bort. Check it out. And, bewaare, the stone men are coming...


Map of Science

This is a great graphic overview of the increasing complexity and interdiscplinary nature of scientific inquiry. (As a side note, it also shows that the social sciences play a much larger role in science than typically acknowledged.) This graph was part of a recent article published in PLoS ONE on 11 March 2009.

Title: Clickstream Data Yields High-Resolution Maps of Science Johan Bollen1*, Herbert Van de Sompel1, Aric Hagberg2#, Luis Bettencourt2,3#, Ryan Chute1#, Marko A. Rodriguez2, Lyudmila Balakireva1

Great Blog: Social Media Today

When people take time to post a comment on this blog, I always take the time to read about their work. Recently, Tom Mandel posted a comment on "Is Foucault a Complexity Scientist?"

One of the blogs on his site is Social Media Today. This is a great site because it is part of the latest trends in internet life. But, it is also an observer of these trends. In short, it is part of the latest movement known as e-science.

As much as I enjoy the web, I find myself in that endless double-bind of participant and researcher. I am fascinated with the web, and yet my researcher side is always asking: What is going on here? Why am I participating in all this? What is this all about? But, no sooner do I ask such questions when I make another click and go: Wow, this is really cool and I've got to tell someone about this new technology or social network, or blog, etc, etc, etc, ugh!

It is because of my double-bind that I really like the blog, Social Media Today. It is a participant in and researcher of the latest trends in information and the forthcoming Web 2.0. Very good stuff for those complexity scientists and sociologists interested in life on the web and where things are going.


Complexity 1001: One More Question

Another overwhelming aspect to sticking my toe into the complexity rapids is the number of new concepts and terms I have encountered (from agent-based modeling to neural networking to fractal geometry, etc.).  So -- in addition to a key/core reference(s) -- what would be the half dozen or so key concepts or terms I would need to master so I can build a foundation in understanding complexity science?  I'm not sure why, but I imagine myself standing on a beach with dozens upon dozens of interesting looking shells -- and while I can picture myself picking up any one of them here and another one or two of them there -- and eventually working my way across all of the shells -- I suspect there would be some shells that are "basic" and thus fundamental to understanding all shells -- and I would appreciate your suggestions here as well.

Dr. Castellani's Reponse

Dear Complexity Challenged, I would start with my complexity science map. Here is why.

The map is conceptual.

Like you, I struggled early on to get a grasp of this field. It is so amazingly interdisciplinary and scattered that it is hard for the beginner (and even expert) to have a true appreciation for what is going on with the field as a whole. After years of struggling to obtain some type of synthesis, I realized that some degree of closure could be obtained if I looked for similarities across the wealth of research taking place. I asked myself, what concepts (be they theoretical or methodological) do all complexity scientists use? And, how do these concepts relate? Also, could I identify the leading scholars associated with these concepts? And, could I highlight one particular sub-concept or area of study with which each of these scholars could be identified? The result was the map.

So, long story short, I would work on mastering the concepts on the map. That will give you an excellent working knowledge and vocabulary sufficient to communicate with any complexity scientist, regardless of their otherwise intractable or incomprehensible research--hee haw!

Complexity 1001: Getting Started

Professor Castellani:  I want to begin a study of complexity -- as it applies to sociology and to issues of healthcare, but I am not sure where to begin.  I've done a bit of googling, read through some of the materials on your site (loved your Complexity Science Map BTW), visited amazon.com -- and at the end of it all, feel a little overwhelmed.

I saw the link for Complexity 1001 and thought I might use it to jump start my learning.

Where would be a good place to start?  What article (book chapter etc.) could you suggest -- something to get my feet wet.  Perhaps from here I could raise a question or two for subsequent discussion, pick up another yet another suggesting resource or two, and go from there?



Dear Complexity Challenged, thanks for becoming part of this blog. I think the best way to "jump in and get your feet wet" is to take a historical macro-level approach and begin with two of the best known reviews of the field.

1. The first is Capra's The Web of Life. While written in 1997, this book still provides the best introductory review of complexity science and its historical roots--in particular, systems science, cybernetics and artificial intelligence and their links to the major themes in complexity science.

2. The second book is Waldrop's Complexity. This is another excellent book because it covers what Capra misses--the historical development of the Santa Fe Institute, the first and most important institute involved in the creation of complexity science and its most cutting-edge research. Almost every major figure in complexity science during the 1980s and 1990s had something to do with Santa Fe. Complexity is a bit journalistic and sensationalist (even gossipy) in style, but it really does give a good historical account of the early years of complexity science.

Most important about The Web of life and Complexity, they introduce you to all the major concepts of complexity science: emergence, self-organization, tipping-points, autopoiesis, self-organizing criticality, computational economics, cellular automata, agent-based modeling, fractals, chaos theory, networks, and so on.

These two books also introduce you to the major players during the 1980s and 1990s: from Holland and Kauffman to Prigogine and Bak to Matarana and Varela.

Once you have a basic sense of the field, you can move to a review of the methods of complexity science. Here is where things become more technical and less macro. You start to move down to the meso and even micro level, exploring specific topics like neural networks, agent-based modeling, the new science of networks, fractals, modeling complex systems, power laws, etc.

But, let's not get into the deep section of the pool too quick. I would get those two books and read them first.

Is Michel Foucault a Complexity Scientist?

In 1999 I wrote an article for Studies in Symbolic Interaction titled, Michel Foucault and Symbolic Interactionism: The Making of a New Theory of Interaction. The article sits at the heart of the theoretical framework (social complexity theory) that Hafferty and I outline in our new book, Sociology and Complexity Science: A New Field of Inquiry. Our theoretical framework, in turn, is part of the SACS Toolkit, which is our new method for modeling complex social systems.

While it may seem odd to some, my journey into complexity science is through the work of Michel Foucault, particularly his later theory of social practice. For me, Foucault’s work has always been about complex social systems and their impact on individuals.

From Madness and Civilization to The Archeology of Knowledge to Discipline and Punish, what are Foucault’s books about? Think about it. At least theoretically and methodologically speaking, they are about complex social systems! Foucault is trying to understand, in post-structural terms, how systems go from one state to another—from one set of self-organizing relations to another. How, for example, does the care of mental disorders, prisoners, deviants, or the self in the west go from a medieval apparatus of care to a modern apparatus of care?

Given this orientation, could we not call Foucault’s work the study of tipping points? Is not Foucault studying how complex social systems evolve over time to become something new, where they suddenly shift from one self-organizing form to another as a function of some type of punctuated equilibrium, some type of major phase shift? Is that not what Foucault’s whole discourse is about, along with the impact these shifting systems have on individuals and their care of self?

Also, could we not call his early work (up to Archeology of Knowledge) a top-down approach to system modeling? Something similar to Luhmann’s view of systems? I mean, is not Foucault, at least early on, trying to understand how systems change without having to call upon some micro-level theory of agency? Something Luhmann and Parsons and others tried to do? Is Foucault not also trying to understand the system within the confines of the system itself?

Then, beginning with Discipline and Punish and his interviews in Power and Knowledge, is not Foucault suddenly grounding his complex systems view in social practice? Suddenly shifting to a bottom-up perspective? Is that not what his methodological shift from archaeology to genealogy is all about? Top-down to bottom-up? A macro to a micro level shift in orientation?

Think about it? How would Foucault sound if he talked about dispositifs and apparatus as complex systems? What if he talked about apparatus which obey their own internal logic as emergent self-organizing systems? What if Foucault talked about his post-structuralism as a way of talking about history as changing dynamic systems that do more than just follow the dialectic? What if he talked about complex social systems that evolve over time along multiple trajectories? Suddenly his idea of systems containing their own resistance (his Nietzschian theory of power) makes more sense: we are talking about the multiplicity of systems, differentiation and feedback loops. And, suddenly his ideas would not seem so unique—at least by today’s knowledge of complexity science. Suddenly his ideas sound less structural and more systems-oriented.

Because this is a blog, I will not blag on too much. So, just consider one of Foucault’s key concepts, the dispositif. For Foucault, this concept forms the field of relations in which his work, up to the end, is situated within.

Foucault states: "What I’m trying to pick out with this term is, firstly, a thoroughly heterogeneous ensemble consisting of discourses, institutions, architectural forms, regulatory decisions, laws, administrative measures, scientific statements, philosophical propositions, moral and philanthropic propositions--in short, the said as much as the unsaid. Such are the
elements of the apparatus [dispositif]. The apparatus [the grid of intelligibility] itself is the system of relations that can be established between these elements. Secondly, what I am trying to identify in this apparatus is precisely the nature of the connections that can exist between these heterogeneous elements (Language, Counter-Memory, Practice, 1980, p. 194)."

As this quote shows, Foucault's work is always about mapping the grid of intelligibility (the dispositif) for some complex system in historical time-—be the system medicine, mental health, the social sciences, criminal justice, psychoanalysis, religion, or government. For Foucault, the dispositif is a system’s self-organizing order of things, its field of organizing practices. But this dispositif is not a totalizing system of relations as in the dialectic. Nor is it something the historian simply uncovers. It is both the interpretive framework that the historian imposes upon the discourses of the past (which is why Foucault often refers to his works as fictions, 1991, p. 33) and the relations that exist between the various discursive and nondiscursive heterogeneous elements making up the field of organizing practices—I mean, does that not sound like 2nd order cybernetics or sociocybernetics? The dispositif is a system of strategies that exist as practice, both on the part of the historian and on the part of the period in question. The dispositif isn’t found within some external structure or within the heads of particular controlling agents. It is within the practice of practice itself. It is fragmented, disjointed and broken, and yet inter-related, unified and organized. It is not a Parsionian system that exists as homeostasis, which then requires us to explain how change happens. It is a changing system where we question how order itself is possible.

Again, this is just a thought. But, it does open up the possibilities for some incredible connections between the last twenty years of sociological inquiry and the new science of complexity. To see a more thorough argument of my point of how Foucault can be used to build a theory of social complexity, see our new book, Sociology and Complexity Science.


I've got a bunch of new geek t-shirts at my CAFE PRESS STORE. Check em out. I particularly like this one and the irony of it.

Factory Wiz


Complexity 1001

Starting today, I will be featuring a new segment on this blog, called Complexity 1001. Like the name sounds, Complexity 1001 will provide an undergraduate (college) level introduction to complexity science and, related, the intersection of complexity science with the social sciences, specifically sociology.

I have asked a few friends who are new to complexity science (a couple profs and a couple students) to post any questions, concerns, or issues they have as they learn about and apply the tools of complexity science.

I also welcome anyone else to post questions they would like answered. You can email me at factory.as@gmail.com or you can post a question in any of the recent Complexity 1001 postings.

Any time I respond to a post, the heading of my post will always be Complexity 1001. This way you can find older postings as the months go by.

Finally, make sure you sign up for a posting feed or all comments feed so you get Complexity 1001 sent directly to your email or whatever place you daily go to see what's happening on the web!

So, let the online course and the postings begin.


Reprise: Intersecting the Study of Social and Complex Networks

Several blogs ago I posted on the need for researchers to do more work intersecting the new science of networks (complexity science) with the sociological literature on social networks, in particular the global network literature. Some sociologists do not see much to be gained from such a merger. For those resistant to the idea or unclear as to what such a merger is about, you need to read Vega-Redondo's Complex Social Networks.

The purpose of this book is to outline, in detail, the avenues of study that emerge from the intersection of the new science of networks and social network analysis.

Rather than going on, I recommend you go to Josep Pujol's excellent review of the book, published at JASSS (Journal of Artificial Societies and Social Simulation)

One note is, however, necessary. Given that my blog caters to social science students and researchers new to complexity science, it is worth mentioning that Vega-Redondo's book primarily makes its case through mathematics. Do not let that scare you away. It is something social scientists have to get used to: complexity science makes extensive use of mathematics to make its arguments. Social scientists are often poorly trained to deal with equation-based modeling. They receive little training outside the study of statistics. We need to get past this hurdle to adopt a much broader and stronger toolset. Having said that, here is one such opportunity to learn something new. Your hard work moving through such a book is worth the effort!


Complexity 5

Okay, so I just posted on Gershenson's blog, COMPLEXES and now I am posting on his recent book, Complexity 5.

I have to admit that this is the exact book I have wanted to write myself. It is a series of overviews (interview style) of leading thinkers in the field of complexity science.

What I particularly like about the book is that it interviews people who other complexity scientists view as TOP NOTCH--rather than the same list of popular people who often get far too much attention. I am particularly excited to see Nigel Gilbert, Paul Cilliers, and Bar-Yam in the list, as well as Melanie Mitchell. There are lots of women in complexity science who have yet to get their dues, and so this is great! (I cannot help making the last point, I am a sociologist.)

Here is the complete list of contributors: Peter M. Allen, Philip W. Anderson, W. Brian Arthur, Yaneer Bar-Yam, Eric Bonabeau, Paul Cilliers, Jim Crutchfield, Bruce Edmonds, Nigel Gilbert, Hermann Haken, Francis Heylighen, Bernardo A. Huberman, Stuart A. Kauffman, Seth Lloyd, Gottfried Mayer-Kress, Melanie Mitchell, Edgar Morin, Mark Newman, Grégoire Nicolis, Jordan B. Pollack, Peter Schuster, Ricard V. Solé, Tamás Vicsek, Stephen Wolfram.

Great Blog: Complexes

I have been following another excellent blog: COMPLEXES.

Complexes is run by Carlos Gershenson, who is a bit of an iconoclast. He has very broad interests in complexity, computer engineering, artificial life, and complexity-based art. He is also the book review editor for Artificial Life and the editor-in-chief of Complexity Digest--the leading compendium of all things complexity on the web.

(As a side note, if you do not have Complexity Digest bookmarked, please do so now. Also, Gottfried Mayer, the founding editor of Complexity Digest and leading systems/complexity scientist, passed away last month.)

Complexes is a fantastic blog because of the range of topics it addresses; and because Gershenson writes in a fresh way, with an insider's insights into various concepts, tools and techniques.

Also, check out his art.

I highly recommend the blog!


Widening the study of global networks

It is time to widen the complexity science vocabulary on global networks. Two rather disparate literature currently exist:

First, there is the new science of networks, and its very specific focus, web science. This literature is dominated by the work of Watts, Newman, Barabasi and scholars in the natural sciences. Web science is a specific focus, examining the world wide web and internet.

Second, there is the globalization literature, and its very specific focus on network society. This literature is dominated by the work of Wallerstein (world systems theory), Manuel Castells (global network society) and John Urry (mobile society and global complexity).

While these two literature are outstanding, not much has been done to bridge them. The closest example from the globalization side is Urry's work in Global Complexity. Related is Wellman's work on web science.

Again, these two literature differ in scholarly background--the first comes from physics and the natural sciences, while the second comes from sociology and political science.

They have lots to say to one another. The global network society literature has a lot to say on the social factors within which global networks are currently situated. The new science of networks has a lot to say about how the structure and dynamics of global networks work--for example, see Barabasi's recent article in nature on mobility.

Disserations, masters theses and funded research await those willing to integrate these two viewpoints in empirically grounded ways.


What are we? Complexity science, complexity theory, complex systems, complex self-organizing systems, etc????

Complexity science has been around long enough for the field to finally settle on a name. Systems science has a clear name, as does cybernetics and agent-based modeling. Complexity science needs similar clarity.

To demonstrate the current lack of clarity, do a basic search on Wikipedia, or examine the various web trackers. It becomes quickly clear that fuzziness and chaos abound--and I do not use these terms in a positive way.

Not having a name is a problem for a new science. It makes it hard for people to know what they are doing.

I strongly recommend complexity science as a name. Here is why

Not Complexity theory: Complexity is more than a theory. In fact, i would like someone to show what complexity theory is? I have yet to see a theory of complexity. I have seen complexity theories about evolution (Kauffman); social systems (Luhmann); organizations (Cilliers). But, I have not seen a complexity theory. No such thing exists.

Not complex self-organizing systems: Of all the possible terms, this one comes close, but it is too cumbersome. Complexity science is definitely the study of complex, self-organizing systems. However, complexity science is broader than just self-organizing systems. It deals with a variety of complex systems. Also, complexity science is cleaner and terse.

Not Chaos theory: While complexity science is indebted to chaos theory, it is something else. It is interested in organized chaos.

Not Agent-based modeling: While complexity science makes use of agent-based modeling, complexity science is more than just method.

Not e-science or web-science: This isn't going to work because the former is too substantively focused and the second is all method and often not systems oriented.

Not Post-systems science or Post-cybernetics: Complexity science is indebted to systems science and cybernetics--these fields are the historical lineage upon which complexity science is based. But, complexity science makes a break with these two fields, turning to a much larger literature to define its theories, methods and substantive problems.

Not Complexity: Complexity science is not just complexity. This term is too wide and ambiguous--we have always had complexity.

Not Computational Complexity: Computational complexity is too focused: it has to do with computational problem solving, not the study of complex systems.

And still more:The other reason complexity science is preferrable is because it separates the field from metaphorical, political and spiritual uses of this new science. A major criticism of complexity science today, particularly in the management literature, is a lack of rigor. Can a car company really be autopoietic? I doubt it. Is emergence some kind of quasi-spiritual mysterious force? If it is, then science might as well stop studying crowd behavior and the standing ovation problem.

Complexity science is a science. Let's call it that.