Epistemology and Unknowns (AGW case study) -- long
A friend, in list-based correspondence about AGW says:
First of all, those of you who are not actual scientists...shut the f*** up an go back to talking about things you know about instead of clouding up the ether with debate over non-essentials. You are making my eyeballs hurt.
I know...I talk to such gentle folks. However, while I think there are deep truths in his position, I think my friend has several levels of confusion on this topic, and I'd like to draw them out.
My friend's argument can be broken into several pieces:
1. First, should we STFU about Global Warming if we're not qualified to have an opinion?
To the extend that opinions are like butts (family blog -- everyone's got one, and they're mostly full of crap), there is certainly a case for shutting up. Furthermore, Patri (and others) recently (Dec/Jan) made the case for specialization in writing, as a way to increase the signal to noise ratio. And I basically agree that if you don't have any experience, you don't know what you're talking about.
On the other hand, AGW is NOT just science. It is also a suite of policy directions, which almost universally are opposed to the libertarian directions that I care about. If one cedes ground to folks who wear lab-coats when they argue policy, then the intelligent move to drive policy is to dress up in lab coats, and then tell people what to do. I do not wish to help create said incentives. Rather, I, as a principled libertarian, have to learn enough about the science to have an opinion of whether I should oppose this or not.
Also, as AGW is a political playing field, certainly the questions of what to DO and what not to do are subject to economic and political analysis. While I do have some sympathy for the idea that folks who don't understand economics should shut up about economics, there's also the political/sociological angle. The idea that 4 trillion euros in play in "green" investing isn't going to get some folks pulling shennanigans in order to get better ROI on that 4T is absurd.
2. Is the scientific question clear?
Is the globe warming? This is a scientific question, to be addressed by reference to temperature measurements, and computer-aided statistical analysis. However, the question is highly misleading in it's simplicity. Is the globe warming is like asking whether people are becoming better off. Looking at things in one way, one time-scale, or in one place (two-parent families in Europe over 50 years, surface station measurements in europe this century) gives different results than looking at things in a different way, time-scale, or place (average wage per hour measured in calories over 4000 years, antarctic ice cores over 100M years).
So first off...Scientists may be better prepared to answer the question, but they are not (much) better prepared to clarify which question is relevant.
3. Who is qualified to offer opinions on that part of the science that is clear?
People qualified to have an opinion are anyone who has experience with the analysis of large hairy datasets or the software used to analyze them. I am suspicious of whether lab-scientists are qualified, because they may have the incorrect opinion that modeling in climate science is similar to their lab-science under controlled conditions. Software folks, Quantitative Science folks, Statistics folks, and even some quantitative social scientists (particularly economists who run models like these) all have the necessary qualifications. Also, lab (or field) techs who take the measurements, or mess with the equipment.
I can personally say, as a software guy, that JUST the comments in the software used at East Anglia show that the model isn't worth as much as the 286 used to run it. If your software is trying to adjust data to get a result, as is OBVIOUS from the code comments, you are not doing analysis, you are doing what is popularly known in the software world as "estimating". For the non-initiated, this means creating an initial value with some degree of rigor, and then changing the numbers until the boss likes them.
I can also say, as a dabbler in economics, that the type of modeling that is done in climate science is awfully similar to the kind of modeling done in macroeconomics. It doesn't appear to predict for shit in either case.
I can also say, as a math-weenie, that this looks an awful lot like a non-linear dynamic system (Chaos, for the hipsters). Non-linear dynamic systems are hell to predict in all but the few cases that we like to talk about in math and science class because we can predict them (ain't gonna happen hell, not very hard hell.).
4. On what grounds do we defer to scientists?
We defer to scientists on the grounds that their information is true.
They are using verifiable data. They are using clear, repeatable processes. Their theory/model predicts experimentally verifiable results. They are using solidly agreed upon theory. The proxy for solidly agreed upon theory is publication of (and citation count of) articles in science journals. Finally, science is assumed to be done in a disinterested fashion. Truth is more important than specific conclusions. All of those things, we don't generally have time to check for ourselves, and it would take a lot of training to do so. In AGW, all 5 reasons to defer to the scientists have broken down.
A. On AGW, the data was not verifiable. It was hidden data, that was not being released. In the face of FOIA the data was not released. Furthermore, climategate emails say conclusively that there was a conspiracy to not release the data (which indicates fear of skeptics poking at it). Furthermore, both Indian and Russian scientists/instrument techs have said that the data that the instruments gave have been manipulated in such a way as to provide the right conclusions. Most recently, the line is that the dog ate the original data. Conclusion: in the case of AGW, you cannot rely on the scientists for data.
B. On AGW, the processes were opaque. First, the software was not released to the world. And it was modeling software of the kind that we know (from experience with Macro) just doesn't work well in general. When the sofware was released through the climategate hack, we discovered that there was a very good reason that the software wasn't released: it sucks. Feed in any data you like (the price of rice in china in the 15th century), and you'll get a hockey stick. Conclusion: in the case of AGW, you cannot rely on the scientists for process.
C. On AGW, the theory and data don't line up ("Hide the decline"). Further, most predictions are effectively non-Popperian. We can't verify. Some of us would say that makes it not science. Conclusion: in the case of AGW, you can't rely on the scientists for experimental verification.
D. On AGW, the peer review process has been corrupted, as per the ClimateGate emails. There was an active conspiracy to keep skeptical voices out of peer review process, and then active claims that "it's not peer reviewed science" against skeptics. The peer review process for climate science is all the way broken. Hence, there can be no supposition that peer-reviewed means good. Conclusion: in the case of AGW, you can' rely on the peer review process to converge upon true theory.
E. On AGW, with all government grants going to climate alarmists, and 4 Trillion(!!!) Euros of green investment funds trying to find ways to make "green" investments more profitable, there is very little chance of disinterested science. Furthermore, those of us who are suspicious of alarmism as per mencken
If you can't get funding for your current studies (or future studies) without coming to pro-AGW conclusions, somehow the AGW conclusions can be teased out of your data.
5. What are the essentials around Global Warming?
As per many previous posts: There are roughly 7 questions in play:
1. Is the globe warming?
2. Is human activity causing it?
3. Is the net effect expected to be bad or good?
4. Can anything meaningful be done (technically) to slow/reverse this?
5. Can anything meaningful be done (politically) to make the changes happen?
6. Is the net cost/benefit to society more positive if we solve global warming, allow global warming, some intermediate, or some alternative?
7. What is the opportunity cost of the results of F?
Right now, the knowns are:
A -- Warming
We have seen no statistically significant warming since 1995.
The cyclical rates of warming and cooling that we have seen have been constant since ~1860.
On the scale of 1860-now, we're warming.
On the scale of 1000-now, it's not at all clear whether we're warming.
Since the last ice age, we're much warmer.
Since dinosaur times, we're much cooler.
B -- Human caused?
Given that there's NO statistical difference between 1860-1880 and 1975-2009, this should be a slam-dunk no. It's not, because the system is so complex that individual components are almost impossible to tease out.
C -- Warming, good or bad?
History says that warming periods are good and cooling is bad, roughly without exception.
D -- Technical solution?
Carbon emission cuts in the politically feasible range (like Kyoto) do effectively nothing to alter the impact.
Technological solutions may help, and may have adverse side effects.
E -- Political possibility?
Not really. Without India, China on board helping, nothing works. Growth is more important than global warming for them. End of story.
F. -- If globe is warming, whether or not due to human activity, what should we do?
Probably, nothing. Canadian/Russian farmland becomes a good investment.
G. Opportunity costs
We should probably, if we look at this as an ideal world-government approach, tax the benefitted Russians, Canadians, Minnesotans, and North Dakotans a little and transfer the gains to the impacted Bangladeshis. In real life, this is a bad idea because it won't get any $ to the Bangladeshis and will create another parasitic government agency.
6. Given:
* we oughtn't just shut up due to technical qualification -- it's a politically motivated question
* the question of AGW isn't particularly clear
* the science is accessible to a moderately large range of folks with quantitative skills,
* the scientific consensus should not be trusted here
* the important questions are not primarily about the science
What is one to do?
Promote skepticism. Don't take someone's word for AGW. Promote careful analysis, and be suspicious of anyone who says stuff is settled. Be especially suspicious of someone asking for your money who says the science is settled. Hope the skepticism doesn't leak into Newton's laws.
aretae.blogspot.com |
