The Flight to the Ephemeral

I'd meant to devote this week’s post to exploring the way that new religious movements so often give shape to emerging ideas and social forms during the decline of civilizations, and to sketch out some of the possibilities for action along those lines as industrial society moves further along its own curve of decline and fall. Still, these essays are part of a broader conversation about the future of today’s world, and now and then some other part of that conversation brings up points relevant to the discussion here.
That’s as much excuse as there is for this week’s detour. A few weeks ago, the P2P Foundation website hosted a piece by Kevin Carson titled When Ephemeralization is Hard to Tell from Catabolic Collapse. Carson’s piece got some attention recently in the peak oil blogosphere, not to mention some pointed and by no means unjustified criticism. It seems to me, though, that there’s a valid point tucked away in Carson’s essay; he’s got it by the wrong end, and it doesn’t imply what he thinks it does, but the point is nonetheless there, and important.

Getting to it, though, requires a certain tolerance for intellectual sloppiness of a kind embarrassingly common in today’s culture. When Carson talks about “the Jared Diamond/John Michael Greer/William Kunstler theory of ‘catabolic collapse,’” for example, it’s hard to escape the conclusion that he simply hasn’t taken the time to learn much about his subject. “Catabolic collapse,” after all, isn’t a generic label for collapse in general; it’s the name for a specific theory about how civilizations fall—those who are interested can download a PDF here—which I developed between 2001 and 2004 and published online in a 2005 essay, and the other two names he cited had nothing to do with it.

Mind you, I would be delighted to hear that Jared Diamond supports the theory of catabolic collapse, but as far as I know, he’s never mentioned it in print, and the modes of collapse he discusses in his book Collapse: How Societies Choose to Fail or Succeed differ significantly from my model. As for the third author, presumably Carson means James Howard Kunstler, the author of The Long Emergency and Too Much Magic—very solid books about the approaching end of the industrial age, though once again based on a different theory of collapse—rather than William Kunstler, the late civil rights lawyer who defended the Chicago Seven back in 1969, and who to the best of my knowledge never discussed the collapse of civilizations at all.

This same somewhat casual relationship to matters of fact pops up elsewhere in Carson’s essay, and leaves his argument rather the worse for wear.  Carson’s claim is that the accelerating breakdown of the existing infrastructure of industrial society isn’t a problem, because that infrastructure either is being replaced, or is sure to be replaced (he is somewhat vague on this distinction), by newer, better and cheaper high-tech systems. What Buckminster Fuller used to call ephemeralization—defined, with Bucky’s usual vagueness, as “doing more with less”—is, in Carson’s view, “one of the most central distinguishing characteristics of our technology,” and guarantees that new infrastructures will be so much less capital-intensive than the old ones that replacing the latter won’t be a problem.

That’s a claim worth considering. The difficulty, though, is that the example he offers—also borrowed from Fuller—actually makes the opposite case.  Replacing a global network of oceanic cables weighing some very large amount with a few dozen communications satellites weighing a few tons each does look, at first glance, like a dramatic step toward ephemeralization, but that impression remains only as long as it takes to ask whether the satellites are replacing those cables all by themselves. Of course they’re not; putting those satellites up, keeping them in orbit, and replacing them requires an entire space program, with all its subsidiary infrastructure; getting signals to and from the satellites requires a great deal more infrastructure. Pile all those launch gantries, mission control centers, satellite dishes, and other pieces of hardware onto the satellite side, and the total weight on that end of the balance starts looking considerably less ephemeral than it did.  Even if you add a couple of old-fashioned freighters on the cable side—that’s the modest technology needed to lay and maintain cables—it’s far from clear that replacing cables with satellites involves any reduction in capital intensity at all.

All this displays one of the more troubling failures of contemporary intellectual culture, an almost physiological inability to think in terms of whole systems. I’ve long since lost count of the number of times I’ve watched card-carrying members of the geekoisie fail to grasp that their monthly charge for internet service isn’t a good measure of the whole cost of the internet, or skid right past the hard economic fact that the long term survival of the internet depends on its ability to pay for itself.  This blindness to whole systems is all the more startling in that the computer revolution itself was made possible by the creation of systems theory and cybernetics in the 1940s and 1950s, and whole-systems analysis is a central feature of both these disciplines.

To watch the current blindness to whole systems in full gaudy flower, glance over any collection of recent chatter about “cloud computing.” What is this thing we’re calling “the cloud?”  Descend from the airy realms of cyber-abstractions into the grubby underworld of hardware, and it’s an archipelago of huge server farms, each of which uses as much electricity as a small city, each of which has a ravenous hunger for spare parts, skilled labor, and many other inputs, and each of which must be connected to all the others by a physical network of linkages that have their own inescapable resource demands. As with Fuller’s satellite analogy, the ephemeralization of one part of the whole system is accomplished at the cost of massive capital outlays and drastic increases in complexity elsewhere in the system.

All this needs to be understood in order to put ephemeralization into its proper context. Still, Carson’s correct to point out that information technologies have allowed the replacement of relatively inefficient infrastructure, in some contexts, with arrangements that are much more efficient. The best known example is the replacement of old-fashioned systems of distribution, with their warehouses, local jobbers, and the rest, with just-in-time ordering systems that allow products, parts, and raw materials to be delivered as they’re needed, where they’re needed. Since this approach eliminates the need to keep warehouses full of spare parts and the like, it’s certainly a way of doing more with less—but the consequences of doing so are considerably less straightforward than they appear at first glance.

To understand how this works, it’s going to be necessary to spend a little time talking about catabolic collapse, the theory referenced earlier. The basis of that theory is the uncontroversial fact that human societies routinely build more infrastructure than they can afford to maintain. During periods of prosperity, societies invest available resources in major projects—temples, fortifications, canal or road systems, space programs, or whatever else happens to appeal to the collective imagination of the age. As infrastructure increases in scale and complexity, the costs of maintenance rise to equal and exceed the available economic surplus; the period of prosperity ends in political and economic failure, and infrastructure falls into ruin as its maintenance costs are no longer paid.

This last stage in the process is catabolic collapse. Since the mismatch between maintenance costs and economic capacity is the driving force behind the cycle, the collapse of excess infrastructure has a silver lining—in fact, two such linings. First, since ruins require minimal maintenance, the economic output formerly used to maintain infrastructure can be redirected to other uses; second, in many cases, the defunct infrastructure can be torn apart and used as raw materials for something more immediately useful, at a cost considerably lower than fresh production of the same raw materials would require. Thus post-Roman cities in Europe’s most recent round of dark ages could salvage stone from  temples, forums, and coliseums to raise walls against barbarian raiders, just as survivors of the collapse of industrial society will likely thank whatever deities they happen to worship that we dug so much metal out of the belly of the earth and piled it up on the surface in easily accessible ruins.

Given a stable resource base, the long-term economic benefits of catabolic collapse are significant enough that a new period of prosperity normally follows the collapse, resulting in another round of infrastructure buildup and a repetition of the same cycle.  The pulse of anabolic expansion and catabolic collapse thus defines, for example, the history of imperial China. The extraordinary stability of China’s traditional system of village agriculture and local-scale manufacturing put a floor under the process, so that each collapse bottomed out at roughly the same level as the last, and after a century or two another anabolic pulse would get under way. In some places along the Great Wall, it’s possible to see the high-water marks of each anabolic phase practically side by side, as each successful dynasty’s repairs and improvements were added onto the original fabric.

Matters are considerably more troublesome if the resource base lacks the permanence of traditional Chinese rice fields and workshops. A society that bases its economy on nonrenewable resources, in particular, has set itself up for a far more devastating collapse. Nonrenewable resource extraction is always subject to the law of diminishing returns; while one resource can usually be substituted by another, that simply means a faster drawdown of still other resources—the replacement of more concentrated metal ores with ever less concentrated substitutes, the usual example cited these days for resource substitution, required exponential increases in energy inputs per ton of metal produced, and thus hastened the depletion of concentrated fossil fuel reserves.

As the usual costs of infrastructure maintenance mount up, as a result, a society that runs its economy on nonrenewable resources also faces rising costs for resource extraction. Eventually those bills can no longer be paid in full, and the usual pattern of political and economic failure ensues. It’s at this point that the real downside of dependence on nonrenewable resources cuts in; the abandonment of excess infrastructure decreases one set of costs, and frees up some resources, but the ongoing depletion of the  nonrenewable resource base continues implacably, so resource costs keep rising. Instead of bottoming out and setting the stage for renewed prosperity, the aftermath of crisis allows only a temporary breathing space, followed by another round of political and economic failure as resource costs continue to climb. This is what drives the stairstep process of crisis, partial recovery, and renewed crisis, ending eventually in total collapse, that appears so often in the annals of dead civilizations.

Though he’s far from clear about it, I suspect that this is what Carson meant to challenge by claiming that the increased efficiencies and reduced capital intensity of ephemeralized technology make worries about catabolic collapse misplaced. He’s quite correct that increased efficiency, “doing more with less,” is a response to the rising spiral of infrastructure maintenance costs that drive catabolic collapse; in fact, it’s quite a common response, historically speaking. There are at least two difficulties with his claim, though. The first is that efficiency is notoriously subject to the law of diminishing returns; the low hanging fruit of efficiency improvement may be easily harvested, but proceeding beyond that involves steadily increasing difficulty and expense, because in the real world—as distinct from science fiction—you can only do so much more with less and less. That much is widely recognized.  Less often remembered  is that increased efficiency has an inescapable correlate that Carson doesn’t mention: reduced resilience.

It’s only fair to point out that Carson comes by his inattention to this detail honestly. It was among the central themes of the career of Buckminster Fuller, whose ideas give Carson’s essay its basic frame. Fuller had a well-earned reputation in the engineering field of his time as “failure-prone,” and a consistent habit of pursuing efficiency at the expense of resilience was arguably the most important reason why. 

The fiasco surrounding Fuller’s 1933 Dymaxion car is a case in point.  One of the car’s many novel features was a center of mass that was extremely high compared to other cars, which combined with an innovative suspension system to give the car an extremely smooth ride. Unfortunately this same feature turned into a lethal liability when a Dymaxion prototype was sideswiped by another vehicle. Then as now, cars on Chicago’s Lake Shore Drive bump into one another quite often, but few of them flip and roll, killing the driver and seriously injuring everyone else on board. That’s what happened in this case, and Chrysler—which had been considering mass production of the Dymaxion car—withdrew from the project at once, having decided that the car wasn’t safe to drive.

The rise and fall of Fuller’s geodesic dome architecture traces the same story in a less grim manner. Those of my readers who were around in the 1960s will recall the way geodesic domes sprang up like mushrooms in those days. By the early 1970s, they were on their way out, for a telling reason. Fuller’s design was extremely efficient in its use of materials, but unless perfectly caulked—and in the real world, there is no such thing as perfect caulking—geodesic domes consistently leaked in the rain. Famed vernacular architect Lloyd Kahn, author of Domebooks 1 and 2, the bibles of the geodesic-dome fad, marked the end of the road with his 1973 sourcebook Shelter, which subjected the flaws of the geodesic dome to unsparing analysis and helped refocus the attention of the nascent appropriate technology scene onto the less efficient but far more resilient technology of shingled roofs. Nowadays geodesic domes are only used in those few applications where their efficiency is more important than their many practical problems.

The unavoidable tradeoff between efficiency and resilience can be understood easily enough by considering an ordinary bridge. All bridges these days have vastly more structural strength than they need in order to support their ordinary load of traffic. This is inefficient, to be sure, but it makes the bridges resilient; they can withstand high winds, unusually heavy loads, deferred maintenance, and other challenges without collapsing. Since the cost of decreased resilience (a collapsed bridge and potential loss of life) is considerably more serious than the cost of decreased efficiency (more tax revenues spent on construction), inefficiency is accepted—and rightly so.

It’s one of the persistent delusions of contemporary computer culture to claim that this equation doesn’t apply once modern information technology enters into the picture. Nassim Taleb’s widely read The Black Swan is chockfull of counterexamples. As he shows, information networks have proven to be as effective at multiplying vulnerabilities as they are at countering them, and can be blindsided by unexpected challenges just as thoroughly as any other system. The 1998 failure of Long Term Capital Management (LTCM), whose publicists insisted that its computer models could not fail during the lifetime of the universe and several more like it, is just one of many cases in point.

The history of any number of failed civilizations offers its own mocking commentary on the insistence that efficiency is always a good thing. In its final years, for instance, the Roman Empire pursued “doing more with less” to a nearly Fulleresque degree, by allowing the manpower of legionary units along the Rhine and Danube frontiers to decline to a fraction of their paper strength. In peace, this saved tax revenues for critical needs elsewhere; when the barbarian invasions began, though, defenses that had held firm for centuries crumpled, and the collapse of the imperial system duly followed.

In this context, there’s a tremendous irony in the label Fuller used for the pursuit of efficiency.  The word “ephemeral,” after all, has a meaning of its own, unrelated to the one Fuller slapped onto it; it derives from the Greek word ephemeron, “that which lasts for only one day,” and its usual synonyms include “temporary,” “transitory,” and “fragile.”  A society dependent on vulnerable satellite networks in place of the robust reliability of oceanic cables, cloud computing in place of the dispersed security of programs and data spread across millions of separate hard drives, just-in-time ordering in place of warehouses ready to fill in any disruptions in the supply chain, and so on, is indeed more ephemeral—that is to say, considerably more fragile than it would otherwise be. 

In a world facing increasingly serious challenges driven by resource depletion, environmental disruption, and all the other unwelcome consequences of attempting limitless growth on a relentlessly finite planet, increasing the fragility of industrial society is also a good way to see to it that it turns out to be temporary and transitory. In that sense, and only in that sense, Carson’s right; ephemeralization is the wave of the future, and it’s even harder to tell it apart from catabolic collapse than he thinks, because ephemeralization is part of the normal process of collapse, not a way to prevent it.

There’s an equal irony to be observed in the way that Carson presents this preparation for collapse as yet another great leap forward on the allegedly endless march of progress. As discussed earlier in this series of posts, the concept of progress has no content of its own; it’s simply the faith-based assumption that the future will be, or must be, or at least ought to be, better than the present; and today’s passionate popular faith in the inevitability and beneficence of progress makes it embarrassingly easy for believers to convince themselves that any change you care to name, however destructive it turns out to be, must be for the best.  As we continue down the familiar trajectory of decline and fall, we can thus expect any number of people to cheer heartily at the progress, so to speak, that we’re making toward the endpoint of that curve.

Not all such cheering will be branded so obviously by another rehash of the weary 20th-century technofantasy of “a world without want,” or that infallible touchstone of the absurd, the insertion of some scrap of Star Trek’s fictional technology in what purports to be a discussion of a future we might actually inhabit. There will no doubt be any number of attempts in the years ahead to insist that our decline is actually an ascent, or the birth pangs of a new and better world, or what have you, and it may well take an unusual degree of clarity to see past the chorus of reassurances, come to terms with the hard realities of our time, and do something constructive about them.