Cambrige Group comments on Chapter 3 The Human Environment System

Please see the attached document.

Introduction to Sustainability Science Seminar

Cambridge group response to Chapter 3 – The Human Environment System

9/28/10

Compiled by Berit Ulrich Berit_Ullrich@hks12.harvard.edu

and Matthew Aruch mia612@mail.harvard.edu

Overall summary:

 The book chapter and readings complimented each other, and it was appreciated that they were not repetitive. A first general impression is that the chapter seems overly anxious to argue for the high complexity that comes into play when you look at the interaction between ecological and social systems. However, what gets sometimes out of view is the question of how you can -- for any specific research task -- set your system boundaries for analysis. What are the criteria to set a CHES system boundary at certain spatial, temporal level and to emphasize a particular subset of environmental and social processes? As there will never be a system delimitation that captures all relevant feed-backs, how can you deal with factors and influences that lie outside of your focus of analysis? By not addressing this issue very explicitly, the reader is increasingly confused to know what he or she can really expect from a CHES perspective and how one would recognize a CHES in ones own research context.

 It remains unclear, for which audience the book and especially this chapter is written. This chapter reads as if it could be written for an undergraduate survey class. But if the book is mainly written for practitioners and graduate students, the point that human-environment systems are complex, can probably be made into two pages at most. The chapter could be halved in length and reorganized. Perhaps a good way to tackle the organization is to use the terms that are used to describe the characteristics of CHES systems in Section 3.1 as "highly complex, often synergistic and rarely deterministic" and make 3 sections that unpack each of these terms with examples. Many of the same examples that have already been discussed could be reorganized under the 3 headings: "highly complex, often synergistic and rarely deterministic". No matter what the terms above need to be unpacked to facilitate clarity. The fact that CHES are complex is also less interesting than how they are complex.

The chapter attempts to answer the question of how with many interesting examples. In general, the chapter felt more like an alternative opening chapter, laying out the territory but not yet getting into much substance. Perhaps the solution is providing more detail and less breadth? For example, are there a small number of mechanisms / interactions that we should know thoroughly? Or are there a smaller number of conclusions that have come out of studying the CHES that we should understand more thoroughly?

 After studying the chapter it still was somehow unclear what the "known" and the "unknowns" are in sustainability science (and in CHE systems in general). Since the field of sustainability science is clearly evolving, it might be useful for this chapter (and perhaps for every chapter) to have a short statement of the "known" and "unknowns" in sustainability science for each chapter topic.

Although the chapter identifies several important aspects of the complexities of the coupled systems, it may be useful to describe how different disciplines that contribute to sustainability science interpret/address these factors.  This may give the reader a better idea of what disciplines/tools/methodologies do a good job at examining or resolving the complexities, and what disciplines/tools/ methodologies are not as well equipped to deal with these factors.

It is not necessary and sometimes not helpful to use that many examples: it feels repetitive and the examples lack salience because they are short and incomplete. Instead the use of a few well-developed examples may be more effective at demonstrating the dimensions of complexity. But on the other hand some of the students considered it really helpful that the chapter is taking the examples developed in the previous chapter (e.g. the Aral Sea debacle, Mayan civilization) and developing these concepts a bit more.

The exhaustive elaboration about the Malthus-Boserup debate might be a little bit outdated. It would be helpful to include the research and findings of the last four decades.  Going back to the note about the audience, this debate may be general information for a graduate level course.

Currently, the text meanders between two extremities: at one level, highly abstract mechanisms are alluded to like path dependency, non-linearities, temporal and spatial autocorrelations, etc. The other level consists of very detailed cases in specific places that illustrate some portions of some of the abstract mechanisms, but for sure cannot be seen as a proof for anything. Perhaps the syndrome perspective from the supplemental reading could provide some sort of middle ground here. They represent a number of ideal-type cases for detrimental human-environment dynamics that could be used as templates for identifying more specific CHES in any new problem to be analyzed. And so it could help to discuss issues like setting system boundaries identification of relevant sub-systems (in the social and the environmental domain), identifying key causal mechanisms

Interesting Points that need clarification:

Considering externalities:

In general externalities happen when we are not considering the full picture. In this case, although we may be modeling the system rather nicely, if we do not take the interaction of the system with the environment, then externalities will show. Considering the effect of externalities and how they can be incorporated in the system (thus having a human-environment interaction) might be a tangible way of approaching this problem. 

 The chapter might benefit from the incorporation of more social theory and the complexities that arise out of humans attempting to understand and contextualize the environment and global change. 

 Consensus on the ecosystem concept:

The authors acknowledge that ecosystem concepts are highly complex.  But at least, some shared understanding about relevant subsystems and essential relationships could be established among different natural science disciplines. However, the situation is radically different for social systems (p. 5, 2nd paragraph). Depending from which discipline you will look at a specific problem, there will be radically different ontological assumptions to start with, irreconcilable core factors identified, largely varying measuring concepts and never ending disputes about the relevant mechanisms at work. If we now try to combine a systemic description of a natural system combined with a conglomerate of largely incongruent social system descriptions, what will result? Is the best you could aspire to identify correlations between macro-variables (e.g. population density being correlated with the number of vegetarians) that do not have much of an underlying causality on which explanation could build?

Additional suggestions:

• Each chapter should begin with a learning objective.  What do they want the reader to understand by the end of the chapter?  Summaries should also be provided at the end.
•  A glossary of terms (that the authors want to generalize for this field) either at the chapter or book level. This glossary could further present the basic shared definitions, possibly acknowledging the contention between the different disciplines.
• At least some reference to the demise of the Maya culture due to overuse of their ecosystems should be included since this is considered one of THE archetype cases for illustrating CHES.
• The role of diversity in social systems is a very hot topic. It is not sufficiently spelled out in the text.
• The text says on p. 1 that human and environmental subsystems are co-adapted among each other. On p. 2 last paragraph the text then says that in the anthropocene the impacts have grown so important that they impact the ecosystems. This caused a logical problem: Are impacts a part of the definition or have they to be considered a problem?
• The introduction of system dynamics (or some other methodology like it) would add an interesting element to the chapter, even so far as it could be a 'straw man' for illustrating the challenges associated with understanding CHES.
• The chapter does not clearly define whether it wants to discuss the "reality" of CHES (i.e. an ontological question) or whether its focus is rather on the means by which we can analytically grasp specific instances of CHESs (a methodological question). If the former was the focus, then there would be a large overlap with chapter 1.2. If its the latter (as the supplemental readings would suggest) then the chapter should be much more explicit on the conceptual tools  that prove helpful in analyzing CHES, i.e. complex system models, agent based modeling, high-level aggregated correlation analysis, intensive case studies, or whatever.
• The CHES examples should also include the positive effect of humans. The Maya culture case may be contrasted with another case, for example the African grassland example developed very concisely in the chapter. Similarly, the Easter Island case is more archetypal and there might be more evidence.

Stylistic and graphic comments:

• Sometimes endnotes are used to refer to a reference and sometimes the author and year are mentioned in the text. What should be the general difference between the endnotes and the reference section?
•  The figures are not self-contained and basically they show something similar, but across different timelines and they explain the CHES slightly different in each. Perhaps it'd be enriching to show a specific example from the text pictorially.
•  Figure 3.2 is too busy to be memorable.
• Figure 3.2 is mentioned on p. 2 but is not scheduled to be placed until p. 5. It would be more helpful for the reader to place the figures near the first mentioning.

I strongly agree with the reactions to the readings expressed above.  I found the chapter, while interesting, to deliver too much abstraction without enough depth to prompt the reader to really work through and understand the concepts presented.  While the purpose of this book and its accompanying reader is to ground people from differing academic backgrounds in the same foundational principles of sustainability science, it does not provide enough concrete terms, definitions and concepts from any one subject to prepare the reader with the confidence to speak knowledgeably about that subject.  I believe that highlighting the key words, phrases and ideas of the chapter and taking the time to define them more rigorously would benefit the readers.  After all, this monograph, as far as I understand, is meant for those with a strong academic background who seek to compliment their knowledge with substantive information from other relevant subjects to sustainability science.

The examples are a pointed example of this flaw in the chapter.  While they provide an overview of how the relationship between human and environmental systems can be complex, they do not go into enough detail to solidify how any of the subjects represented by the field of sustainability science can inform the understanding of the example, nor how several subjects can communicate about it.  Additionally, I found that the examples and discussions went into very little detail about the human subsystems at play in a CHES.  I saw a large amount of time spent on the changes in environmental subsystems that took place as a result of human actions, but found substantive discussion about the workings and relevance of human subsystems (economics, policy, etc) lacking.

On a different note, I did enjoy the interplay between the chapter and articles from the Reader.  I thought that Reader introduced a new and important way of looking at CHESs, displaying the application of concepts discussed in Chapter 3.  Both the models and "syndromes" discussed presented true attempts at tangibly understanding and explaining human-environment system interactions, which helped me form a stronger and more comprehensive picture of the main points and concepts of this chapter.

Being a member of the Cambridge group, I originally had many of the same reactions as my fellow students about the book chapter and even voiced these ideas for our compiled comments.  In particular, I found the chapter overly vague, choked with half-explained examples, and not lucid in describing the “knows” and “unknowns” of CHESs.  However, after re-reading the chapter, thinking more about these issues, and re-considering the Cambridge group’s compiled comments, I am beginning to be able to articulate a subtle but pervasive concern with this chapter—and many of the chapters—that I was unable to verbalize before.

Below is one of the comments that I have continually revisited:

“The text meanders between two extremities: at one level, highly abstract mechanisms are alluded to like path dependency, non-linearities, temporal and spatial autocorrelations, etc. The other level consists of very detailed cases in specific places that illustrate some portions of some of the abstract mechanisms, but for sure cannot be seen as a proof for anything. Perhaps the syndrome perspective from the supplemental reading could provide some sort of middle ground here. They represent a number of ideal-type cases for detrimental human-environment dynamics that could be used as templates for identifying more specific CHES in any new problem to be analyzed. And so it could help to discuss issues like setting system boundaries identification of relevant sub-systems (in the social and the environmental domain), identifying key causal mechanisms”

I too initially found the chapter’s hormonal oscillation between abstract, generalized, global concepts and actual, applied, specific, local examples to be quite vexing.  Originally I thought my consternation was because this approach prevented the chapter from conveying its central points.  However, I now believe that this chapter’s organization (perhaps inadvertently) conveys one of the central tensions of sustainability science (and academia in general): the conflict between generalized theory and applications.

As academics involved in the definition and development of a discipline, we have a strong penchant for wanting to find the broad generalizations that apply to all situations.  Certainly focusing on generalized theory helps us establish a common language and lens through which to view and discuss phenomena.  But we also spend a great deal of time on generalizeable principles so that we can create a set of generalized (but customizeable) solutions that we can apply to the problems that we are attempting to solve.  I believe that this is the appeal of the “syndromes” approach in the Ludeke, Petschel-Held and Schellnhuber article—we identify a syndrome X that has an associated set of therapies Y; therefore, if we see syndrome X in the “real world”, we know we can take one (or more) of the therapies Y and (with slight context-specific modifications) apply it to the syndrome and resolve the problem.

However, coupled systems may NOT be able to be generalized any more specifically than the author currently describes, and THIS might be the real point of frustration for all of us.  We do not actually observe CHESs; we cannot see an “unsustainable urban growth” system or a “governance system that promotes unsustainable demand”.  What we observe are INDICATORS--manifestations of the INTERACTIONS of different systems.  However, if indicators are the product of interactions of systems, and if all systems are actually aggregated indicators of other subsystems, then indicators may actually be so specific to the particular systems of subsystems that exist in that particular context it becomes IMPOSSIBLE to generalize at a more specific level.  Thus the problem may not be with the author’s organization of the material, but actually a product of the fact that CHESs do NOT have any commonalities beyond the “abstract” concepts of “linkages”, “special and temporal levels and effects”, “diversity”, “path dependence”, and “non-linearities”.

If this is indeed the case, it is exceedingly annoying for us, as we cannot satisfy our desire to develop a set of “(almost) ready-made” solutions or prescriptions to CHES problems.  It means that sustainability problems are so cite- and context-specific that each “therapy” HAS to be created specifically for that cite and context.  We therefore many NOT be able to identify system “templates” or speak any more specifically about “relevant sub-systems” or “key causal mechanisms” without discussing a PARTICULAR example.

However, part of me certainly wants to believe that sustainability science will enable us to classify a set of archetypical CHESs and develop a corresponding set of standard interventions to CHES problems—it would make the policy process significantly easier and less costly.  Therefore I am open to being convinced otherwise—so please feel free to respond!

Although the chapter addresses a quite interesting topic of how to couple human and environment systems, the style and structure in which is written makes it difficult to follow. There are too many unconnected ideas and examples explained in very little detailed. It strikes me as a quite different chapter than the previous one where the text was engaging and clear. A way to improve this chapter is to enhance the cohesion within itself and between the previous chapters. Use less technical language for making it more accessible to people unfamiliar with the field. Incorporate the ideas introduced in the presentations, like defining boundaries, integrating disciplines, and the battery pack analogy.  Thanks very much to the presenters for making CHES more clear!

Previously Matthew Aruch wrote:

Please see the attached document.

CHES - ontological vs. methodological:

The Chapter seems to present CHES, considering humans and the environment as separate, as an ontological concept (i.e. a statement of being - humans should be thought of as separate from the environment because it is intrinsically so), rather than a methodological concept (i.e. humans are considered separate from the environment because it is methodologically useful from an interdisciplinary perspective).  However, the presentations on Monday seemed to suggest that this was open to debate.  Clarifying this in the chapter would be very useful

Previously Juan Pablo Giraldo wrote:

Although the chapter addresses a quite interesting topic of how to couple human and environment systems, the style and structure in which is written makes it difficult to follow. There are too many unconnected ideas and examples explained in very little detailed. It strikes me as a quite different chapter than the previous one where the text was engaging and clear. A way to improve this chapter is to enhance the cohesion within itself and between the previous chapters. Use less technical language for making it more accessible to people unfamiliar with the field. Incorporate the ideas introduced in the presentations, like defining boundaries, integrating disciplines, and the battery pack analogy.  Thanks very much to the presenters for making CHES more clear!

Previously Matthew Aruch wrote:

Please see the attached document.