Topic 2: How do we integrate results produced by distinct disciplines? What do you do when research leads to different outcomes? Are we able to integrate disciplines without producing a hierarchy among them?

A good start to integrating results across disciplines might be to first persuade those within each of our respective disciplines to acknowledge the ways in which human-environment interactions affect their own results. Ideally, economists and political scientists would make it regular practice to consider environmental feedback effects in interpreting their results, and natural scientists would make a habit of considering how their results might change in light of social concerns. It would then perhaps be less difficult to reconcile the differing results implied by distinct disciplinary methods.

I recently sat in on a lecture about the ongoing debate within development economics with respect to randomized control trials (RCTs). Though I’ve heard both sides of this debate several times before, this time I found myself thinking about the arguments in terms of our discussion of complex human-environment systems. The gist of the debate is that while RCTs provide a high-degree of internal validity, their external validity is low and they often fail to consider “general equilibrium” effects that might result from scaling up the intervention in question. It occurred to me that if we broaden the concept of “general equilibrium effects” to include long-term environmental feedback effects, then really neither side of the development debate has taken this big picture view.

On average, neither RCTs nor large cross-country econometric analyses do justice to the complexity of human-environment systems. Aside from disregarding additional outcomes of interest (like contributions to worldwide carbon emissions), such omissions could even invalidate the very outcome that the paper was intended to study. For example, if an intervention degrades local natural resources, then in the long-run there may be negative externalities for the very same population that was intended to benefit. Similarly, conservation programs that fail to consider negative impacts on local livelihoods may be expected to increase economic pressures on those resources in the future.

I wonder if other disciplines are currently engaged in similar methodological debates, which could be informed by insights from our discussion of CHES. I am still uncertain how such complexity could be fully captured in a single study (and am very much looking forward to the chapter dealing with methodology). However, a promising first step would be to at least recognize the exclusion of human-environment interactions as a relevant critique. Even if it is not the central focus of a paper, when the authors are discussing implications and extensions, a discussion of human-environment interactions should be standard.

On the more general question of how sustainability science integrates knowledge across topics: the "battery pack" approach is very much about transcending other disciplines. Sustainability science is the field that looks at emergent properties, and that looks at theories (not just problems) that span across several rows/columns in the matrix of week one. 

There is, though, an alternative approach. Instead of overlaying other disciplines, it could steal from them - taking what it needs from economics, biology, etc etc.  A seminar structured in this second way would not just look at theories that cut across disciplines, but at what each discipline has to contribute. It would say: ok ecologists and geographers, here are the ten things about sustainability that get covered in economics 101. And economists, here are the things you need to know from physics. And so on.

Then, when it comes to particular problems, you don't just have a battery pack but a choice of tools, or at least a knowledge of when other tools are needed.

Would this be a better approach? Some thoughts:

1 - Coming (roughly) from economics, I'd say there is a list of insights that are almost a minimum requirement to talk credibly about pollution - including the background of externalities and public goods, common property, coase, pigouvian taxes, and how economists think about discounting and welfare aggregation. Simple stuff and I suspect most people know it. But I have not a clue - not the faintest idea - what the equivalent list is for other fields, and would love to find out.

2 - An analogy: we are faced with not understanding each other's language, and instead of trying to learn a bit of French, bit of Spanish, we are trying to invent Esperanto. That allows us to talk to each other, but doesn't particularly give insights into how different cultures think about things.

3 - Of course, everyone would be bored in at least one of the classes, and the broader their knowledge at the start, the less useful it would be for them... I'd probably find every class useful!

The questions, I think, are whether it is (a) possible, and (b) desirable to only focus on battery packs - or whether we need to do the other approach as well.