The Perils Of Efficiency

Largest Ever Study Finds GM Crops "Harm Wildlife" The world's biggest scientific experiment into the environmental impact of genetically-modified crops, conducted on British farms, has shown that GM rapeseed and sugar beet are more harmful to wildlife than conventionally grown plants. ... Scientists unveiling the results at the Science Centre in London said some insect groups, such as bees in beet crops and butterflies in beet and spring rape, were recorded more frequently in and around conventional crops because there were more weeds to provide food and cover. ... Researchers stressed that the differences they found were not a direct result of the way in which the crops had been genetically modified. They arose because the GM crops gave farmers taking part in the trials new options for weed control.

The upshot of this seems to be that GM herbicide resistance works too well. The ideal of a modern farm is to have all crops of even quality, with no other organisms (such as pests and weeds) damaging or competing with the target crop species. Without modern agricultural technology, this ideal could only be roughly approximated. But advances -- most recently the development of GM -- have pushed us closer and closer.

Any ideal is a simplification and an abstraction. The real world is messy and complex. And to some degree I think the world depends on its messy complexity. Too perfect a realization of an ideal can undermine a system. For another crop-based example, take the genetic uniformity promoted by the use of high-tech hybrid and GM seed varieties. These even out the variable quality of more "natural" plants, pushing us closer to having the Form of the plant visible here in the cave. But that very variability becomes crucial when the crops are subject to stresses from climate or pests and diseases, and need the raw material for evolution. Perhaps the problem could be eliminated if we could realize our ideals across the entire universe, so that these idealized systems wouldn't be upset by their interface with the remaining messy parts of the world. But that degree of control is not in our power. So our idealized systems depend on their residual messiness to sustain themselves.

The model of genetic diversity doesn't entirely map on to the "weeds are good for wildlife" conclusion of the article I quoted. The article gives no indication that the wildlife are good for the farmer, presenting them instead as a worthwhile value in addition to farm productivity. One could retort "well, good riddance to that wildlife. It shouldn't be in the field in the first place." But it's also plausible that there is in fact such a connection (either directly or in a larger landscape health sense), in addition to the fact that one could argue that "farm productivity" and "wildlife" are both contributors to the larger goal of a good society.

This post is not an all-out attack on creating idealized systems (I don't think it's even a definitive statement that, on the basis of the study quoted, GM herbicide resistance is too idealized). Some degree of idealization is how people work, how we interact with our surroundings. It's often a very productive method. The key is to be cautious about going too far, and falling victim to the hubris of thinking we can get everything under control.