In the song by Harold Arlen and Johnny Mercer, that old black magic is, of course, love -- powerful and unpredictable, common yet mysterious, and completely transformative. Thus, it is frequently overlooked until it happens to pounce on one. Technology is not dissimilar.

This song has crossed my mind a number of times recently as I have followed scientists discussing how the Colorado River watershed, and thus water resources in the American West, will likely be affected by global climate change. One of their major points is that under most scientific scenarios Phoenix, Las Vegas, and other desert cities will inevitably run out of water.

By their analysis, this will occur as precipitation patterns shift, the River flow shrinks and water appropriation law takes over - in this case, given that Arizona's and Nevada's water rights to the Colorado are junior to California's, California gets all the water that's left. So from their - scientific -- perspective, Phoenix will wither in the desert.

But from an engineering perspective, this is simply wrong, for water availability almost anywhere is a pricepoint issue, not an absolute availability issue. A rise in price, reflecting increased scarcity or demand shifts, puts more technologies into play. In the short term, one gets technological substitution as existing but more expensive technologies become economic with price increases.

Over the longer term, however, one tends to get unpredictable and fundamental improvements in technological systems. Thus, for example, if the flow in the Colorado River shrinks, there will be a point - more expensive, but not technologically infeasible - when Arizona can build desalinization plants, either in California or Mexico, and treat all the water it needs.

Expensive? Of course. Political quagmire? Water in the West always is. But will we "run out of water" as some scientists fear? Not unless we choose to.

Global climate change plays the same tune. A major theme of that debate is the need to reduce fossil fuel use to reduce carbon dioxide emissions, bolstered by the 200 year projections that are routinely used in climate change debates.

But such projections are far more speculative than they appear and than their adherents frequently claim because they systemically underestimate the unpredictability and power of technological evolution. For one thing, the focus on emissions presupposes a technological stasis that is historically questionable. For example, recent work at Carnegie Mellon University and elsewhere, including development of pilot scale technology, suggests that extracting carbon dioxide from the ambient atmosphere, liquefying it, and injecting it into geologic formations, is both economically and energetically feasible.

This research is in prepublication stage, but if it, or any similar technology, is developed, it obsoletes our current understanding of, and approaches to, the "climate change problem." The question shifts from fossil fuel use and emissions, to the much more profound one of what atmospheric chemistry, and dynamics, we as a species choose to have - and who gets to make that choice.

This case also illustrates another fundamental weakness in current approaches, the tendency to project current trends into the future without assuming adjustment by other components of these complex human/built/natural integrated systems. Thus, it is quite simply wrong to view computer projections of states well into the future as somehow "real," for one thing we know for sure is that current rate of social and technological change makes them worthwhile warnings at best. Valuable thought experiments, perhaps; probable in any sense, no.

In part, we cling to such projections, wrong though we know them to be, because we fundamentally don't understand technology and technological systems. Moreover, we have a very human tendency to privilege the present, resulting in strong aversion to even the idea of technological evolution.

Additionally, the agenda of proponents of particular policies may only in part be one of environmental quality; in many cases, it includes the more implicit social engineering goal of reducing consumption and restructuring capitalism, especially in America. Thus, one concern expressed by environmentalists about ambient atmospheric carbon dioxide capture technologies is that because they enable continued use of fossil fuels they don’t adequately constrain economic growth.

Understanding the power of technological evolution does not mean that technology will answer all our problems, nor justify unthinking technological optimism. But our abject failure to appreciate the power of technology does indicate that the current intellectual frameworks and tools we bring to understanding our anthropogenic earth are badly flawed, and that we do not yet know how to take ourselves, and our world, seriously.

--------

Brad Allenby is professor of civil and environmental engineering at Arizona State University, a fellow at the University of Virginia's Darden Graduate School of Business, and previously was AT&T's vice president of environment, health, and safety.