ES -- DX/CL -- isee -- cboe put/call -- specialist/public short ratio -- trinq -- trin -- aaii bull ratio -- abx -- cmbx -- cdx -- vxo p&f -- SPX volatility curve -- VIX:VXO skew -- commodity screen -- cot -- conference board

Wednesday, June 08, 2005


chaos and complexity

via reason, word that national academy of sciences has issued a statement saying that anthropogenic climate change is "real" and must be acted on immediately. ronald bailey sums up their view:

By adding greenhouse gases to the atmosphere (chiefly carbon dioxide from burning fossil fuels), humanity is increasing global temperatures. How much? Uncertain. How dangerous? Uncertain. How best to handle it? Uncertain.
but i have a more fundamental question regarding the "reality" of all this.

let's say i accept the evidence as it is believed to be known:

Carbon dioxide levels have increased from 280 ppm in 1750 to over 375 ppm today -- higher than any previous levels that can be reliably measured (i.e. in the last 420,000 years). Increasing greenhouse gases are causing temperatures to rise; the Earth's surface warmed by approximately 0.6 centigrade degrees over the twentieth century.
what have these observations got to do with this conclusion?

The Intergovernmental Panel on Climate Change (IPCC) projected that the average global surface temperatures will continue to increase to between 1.4 centigrade degrees and 5.8 centigrade degrees above 1990 levels, by 2100.
putting aside the moral argument, isn't it quite clear that modeling an inconceivably complex, massive and chaotic system like earth is essentially impossible? that any model we can construct will be such a reduced and incomplete version of the reality as to be essentially purposeless for prediction, except as a vehicle for self-delusion? and that the models in existence have never even approached (and can never approach) anything like induction by enumeration -- the fundamental principle of science -- because they cannot test against reality repetitively to gain confidence?

it seems to me that the irreducible complexity of the system will always defy meaningful analysis, despite what adherents of scientism and the cult of techne might like to believe. our models -- such as they are, being hopelessly reductive -- are constructed essentially by backtesting some broad ideas and adjusting parameters to fit the data we have collected on the recent past -- itself a sample size too small to be significant -- and in any case without anything that an appropriately honest scientist would call a well-understood mechanism.

under such circumstances, any prediction our integrated assessment models of the global environment eject on a hundred-year scale then is little better than a coin flip -- for the same reasons that mathematical models backtested to fit the stock market invariably lose money. the reality is itself unpredictable -- the actual open system is both complex and chaotic. observation of past behavior yields no mechanism by which long-term future performance can be even hinted at.

so why should these predictions be treated seriously?

i don't think they can be, frankly. more co2 does not mean higher temperatures in the same way as pushing down on one end of a seesaw raises the other end. that is only assumed by some to be true because of a presumed correlation between atmospheric greenhouse gas content and global temperatures which may in fact be specious.

but that's all besides the point, anyway, isn't it? the earth isn't a seesaw. you apply an impulse A under conditions A' and observe outcome X. the you present impulse A under conditions A' and observe outcome Q. do it again and get outcome Z. the system is fundamentally not predictable; it is complex and chaotic, and people would do well to stop pretending that they have any idea what effect any disturbance -- up, down or same -- will have on it. it isn't a matter of "degrees of uncertainty" at all. the system does not present reproducible outcomes over time under any circumstances, and may very well give opposite outcomes on identical impulses. more co2 might mean warming -- or it might mean cooling -- but which way and to what degree? there is no single answer, and can never be one. the system is both complex and chaotic.

there are circumstances under which some complex systems will behave predictably for short time periods. certain impulses do set up regular responses. but only a diminishingly small minority, as it turns out. and, yet more vexing, the earth is not only complex but comprised of a nearly-infinite number of interacting chaotic systems.

this topic of climate change in the context of complexity and chaos is *vastly* more sophisticated than most ordinary people are capable of treating it, and goes to the heart of why the scientismic philosophy of the world as a clockwork that needs only to be revealed is profoundly wrong -- and why science is, in many places, at the limits of what can be known through it, leaving charlatans and believers in scientism left to surpass those limits with myth and fiction.

cary neeper provides an introduction to complexity and chaos which can function as a primer to a very difficult, counterintuitive topic.

Labels: ,

Here is a good example of the oversimplification of climate change. I expect no more from McPaper.

------ ------- ------
Wow. Something we can completely agree on.

Check out for an interesting take on the importance of global warming, relative to more pressing issues. The global warming debate seems riddled with squishy science, but more importantly, diverts attention and resources (Kyoto's projected costs are in the hunreds of billions) away from real, current problems.

Nice post.

------ ------- ------

Post a Comment

Hide comments

This page is powered by Blogger. Isn't yours?