The Control and the Scientific Enlightenment

Raindance.jpgChris Bertram of Crooked Timber makes the following statement:

Suppose there are two possible states of the world, S1 and S2, and we don?t know which of the two states the world is in. An event E occurs which is consistent with the world being in either S1 or S2, but is more likely in S1 than it is in S2. We should surely say that, given E, the world is more likely to be in S1 than in S2, and that to that extent E (though consistent with both possible states) is evidence for the world?s being in S1.

Such evidence isn?t, of course, conclusive. After all, by hypothesis, E is consistent with both possible states. But evidence doesn?t need to be conclusive evidence to count as evidence.

Chris states in the comments that his question is not primarily about global warming but rather about whether bits of information can serve as evidence to give people reason to revise their beliefs. From what I can surmise, Chris is using the following set of parameters...

S1: the world is warming due to human influence

S2: the world is not warming due to human influence

E: last summer, it was really hot in Europe

...and implies that since E is a piece of evidence for S1, the world is more likely to be in S1. That seems to work out very well. Each little piece of evidence should be used to raise our confidence of the state of the world to the degree that that piece of evidence can do so, even if it does not do so conclusively.

Yet, thinking about things a bit more, I changed the framework slightly.

S1: the world is warming due to human influence

S2: the world is not warming due to human influence

E: last summer, it was quite cool in the northeastern United States

Now it might be argued that this bit of information 'nullifies' Chris's bit of information because it seems to imply that the world is actually in S2, but does not necessarily nullify Chris?s assertion that bits of information can be used to make conclusions about the state of the world, since my E is simply another piece of evidence adding to the already existing E that Chris obtained. But then I thought of another example:

S1: full moon causes violent behavior in humans

S2: full moon does not cause violent behavior in humans

E: An emergency room physician admits 5 stab victims during a full moon night

What does E imply in this case? Is it more consistent with S1 or S2? Does it help acknowledge one view of the world or another? Or how about:

S1: Jonathan Wilde can cause rain by doing raindance

S2: Jonathan Wilde cannot cause rain by doing raindance

E: Jonathan Wilde does raindance and rain follows

Or how about these two examples:

S1: frogs come from muddy soil

S2: frogs do not come from muddy soil

E: large numbers of frogs appear in muddy soil after River Nile floods


S1: rotting meat turns into fruit flies

S2: rotting meat does not turn into fruit flies

E: fruit flies appear around carcasses hung in the market

E seems to suggest that S1 is reality and S2 is a myth in both of these examples. Huh? What?s going on?

What are we to make of all this? It is obvious that the world is full of information, and pieces of evidence may not always lead to the correct conclusions. Man's understanding of nature can often be misleading and many times, be completely at odds with reality.

When people think of the scientific method, they usually think of it the way I was taught in school ? make observations of the world around you, come up with a hypothesis, carry out actions to test the hypothesis, analyze the data, and make conclusions about that original hypothesis. Although it may seem systematic and rigid, it is quite evident man has been doing this sort of thing since leaving the caves. Yet, it was this exact method that led some men to believe that performing raindances could cause rain to fall, that rotting meat turns into fruit flies, that sacrificing virgins atop holy shrines can bring peace upon the land, and that bloodletting can cure disease. This limited empirical approach gave man a very flawed understanding of objective reality and resulted in a worldview haunted by sadistic demons, themselves emotionally juvenile, at times cruel, spiteful, and capriciousness. Some people even gave up altogther on man's ability to understand the universe, forever resigned to his imperfection in empiricism.

In the evolution of scientific thought, there was one gigantic leap of methodology that did away with much of the haziness of man's understanding of the universe. This approach revolutionized the way knowledge was obtained and cleared away much of the murky floatsam resulting from our observations. It was the control.

The control is the part of the experiment in which the independent variable, i.e., the variable under study, is non-existent. If there is a difference in outcomes of the two groups under study, then this difference can be attributed to the differences in inputs. In the case of a controlled study, the only difference in inputs is the independent variable. As a result, the control is no ordinary measure. A true control makes what was once impossible in the prior demon-haunted world possible : it allows correlation to imply causation. And in doing so, it vanquishes the demons.

One of the first controlled experiments was performed by an Italian physician named Francesco Redi who actually put lids on jars containing rotting meat to create a control, and his observations led him to rightly conclude that the theory of spontaneous generation was bunk. Fruit flies do not come from rotting meat; rather, fruit flies come from other fruit flies. And controlled experiments since his time have disproven raindancing as a means to cause rain, the notion that the full moon results in more emergency room trauma admissions, heavy objects falling faster than lighter objects, muddy soil turning into frogs, bloodletting curing disease, and prescriptions for complete bedrest after surgery, among many other things.

Without the control, what remains is anecdotal evidence. Anecdotal evidence ruled the world for thousands of years and gave rise to a faulty understanding of the world. It was severely limited in its ability to explain nature. Only when it was abandoned in favor of controlled evidence did science become part of the Enlightenment and finally serve to unlock the mysteries of the universe. The modern day advances in science could never have come about if the control had not become an essential part of the scientific method.

So when Chris espouses anecdotal evidence that Europe?s hot summer ?raises our confidence? that global warming is occurring, he is going back in time over 300 years when empiricism was little more than mysticism and faith. Step across the chasm from that dark side of mystery into the methods of The Enlightenment, Chris Bertram. We are waiting for you.

Share this

Very nice explanation of

Very nice explanation of important scientific point. The need for "a control group" is why most economics is not, and almost can not be, THAT kind of science. Note on global warming the inability of weather models to show what caused "the little ice age" around 1200-1400 and such variations. Note that LAST summer, there were floods in Europe.

I actually DO believe the increase in CO2 (what is decreasing? almost everything else; but slightly, cause there's not so much CO2) is likely to increase the fluctuations: hotter hots, colder colds, wetter wets, drier drys -- as well as the moderate inverses like colder summers. But I accept this is more belief, than science.

So a rev-neutral increase of the gas tax, matched by a reduction in VAT, would prolly be a bit better.

It is much harder to heat

It is much harder to heat wet air than dry air.

The warmer it is, the more water is absorbed.
The colder it is, the less water us absorbed.

Thus, cold air is easier to warm because it is dry.

Warm air from the tropics and subtropics are not going to warm appreciably because of energy transfer (it takes orders of magnitude more energy to raise tropical air a degree than siberian air).

Hence, it cannot be the case that increased CO2 will cause more "extremes", since water buffers out extremes. A cooling climate would cause more heat extremes (less water in the air) than a warming one. A warming trend will tend to have the world as Florida (even in Scandinavia) before it has the world as the Sahara. A cooling trend, though, would increase temperature inequality.

And it is temperature inequality that drives all weather in the world. So is it more, or less likely that with a warming trend we are to see more "extreme weather"?

Most of this is the thesis of "The Satanic Gases" by Michaels and Balling, from the Cato institute. Basic stuff, but compelling (especially in light of the evidence- troposphere warming with stratosphere cooling).

Jonathan's observation is an

Jonathan's observation is an excellent one, and goes a long way to explaining why so many people are made uneasy by claims that the science of climate change is "settled"--admittedly, this claim is more often made by policymakers and activists than by, say, IPCC WGI. The claim is usually followed by the assertion that we should all stop debating and get on with "doing something", even if "something" may be entirely ineffectual or even counterproductive.

Very strange to see such

Very strange to see such championing of Enlightment rationalism and denigration of tacit knowledge on a blog with an Austrian economics name! Surely one could extrapolate your argument to suggest that only the rational planner can ensure a scientific allocation of resources, while messy market-based practices are exactly the sort of messy groping about that kept us back in the Dark Ages for so long?

Or to put it another way; however much you might want to, you can't quarantine the critique of planning to the economic sphere; it's a comprehensive epistemology.

Very strange to see such

Very strange to see such championing of Enlightment rationalism and denigration of tacit knowledge on a blog with an Austrian economics name! Surely one could extrapolate your argument to suggest that only the rational planner can ensure a scientific allocation of resources, while messy market-based practices are exactly the sort of messy groping about that kept us back in the Dark Ages for so long?

Absolutely not. It is this very argument that lends credence to the Austrian critique of empiricism in economics - the lack of feasible controlled experiments.

Perhaps the natural world is

Perhaps the natural world is knowable in the sense that natural rules do not change either with purpose or at random, while human designs, desires, wants, and ends DO change on a regular basis (again, either by design or at random, whichever explanation one prefers).

Thus one can go about studying the universe in a rational manner and plan experiments and, relying upon unchanging constants of nature (physical laws), can predict with great accuracy the course of inanimate objects and how things will react. This is something inherently impossible to do with people, or anything social. There are no constant relations between individuals, social units, institutions, or otherwise that gives an analog to the methods of the physical sciences (and thus would allow expert planning).

And furthermore, how can one

And furthermore, how can one possibly liken belief about global warming to 'tacit knowledge'?

Surely it cannot be suggested that the concept of global warming is on par with Pythagoras' Theorem, for example (something that is knowable without recourse to empirical verification)...

one could probably also

one could probably also apply the same logic to the Iraqi WMD debate:

S1: Saddam Hussein has WMD.
S2: Saddam Hussein does not have WMD.
E: The UN Inspectors cannot account for the WMD the United States gave to Saddam during the 80's.

Spoonie, the US didn't give

Spoonie, the US didn't give any WMD to Saddam in the 80s.

The US had practically no trade with Saddam in the 80s. Brazil sold more weaponry (of a conventional sort, at the every least) than the US did in the 80s.

This misconception keeps popping up, like a whack-a-mole game.

interesting point spoonie

interesting point spoonie love, because if your analogy is appropriate then most conservatives would be inconsistant in their dealings with the two issues, as I think is obvious to anyone.

On a perhaps related note though, I've always despised the "we don't know if it's global warming" argument because, while not quite as bad as the "evolution is just a theory" arguments, it is often used as if it's a good counterargument to "we should be concerned about global warming."

We can write off any temperature change as as just a shift in global temperature patterns if we want to. The problem with such a method is that since we don't have old enough to data to even grasp what a cycle might look like, broadly (according to some) we couldn't hope to understand what a human-made change might look like. This is an argument I use on theists all the time: what would it take to convince you that your beliefs are false- i.e. that global warming is real?

In this situation, though, Pascal's wager works the other way. Supposing that "humans are increasing global warming" is a basically untestable hypothesis, presently, what would be the most rational decision?
a. go on about our business
b. providing such a controlled experiment by enforcing an international agreement reducing certain emissions, and testing that reduction's effect on global temperature. The montreal protocol is a good example of such a succesful agreement.

By the way, I should remind people that, hypothetically, global warming seems very intuitive. If you drastically increase the presence of certain chemicals in the atmosphere, it seems intuitive that you would produce some tangible effects on earth. Hence, it's not really akin to a "full moon" type of argument (though oddly enough, there is evidence supporting such a hypothesis.)

-Matt

My Internet handle,

My Internet handle, Spallenzani, is named after Abbe Lazzaro Spallenzani, who played a significant role in the history of microbiology as the first person in history to perform a controlled experiment, at least according to this link.

Hey Matt, glad to see you

Hey Matt, glad to see you found the comment threads. :)

I'm not sure if your theist argument proves much other than the fact that Karl Popper wouldn't consider theism a science. Do you think someone like Kierkegaard cares if his faith is falsifiable?

I'm far from knowledgeble on questions relating to global warming, but your Pascal's wager argument doesn't convince. Even if we assume that global warming is untestable given our current data, that in itself is not a justification for extremely costly experiments like Kyoto. Perhaps the Montreal Protocol is on a much smaller scale; I've never heard of it before.

As for the intuitiveness of global warming, keep in mind that global cooling was also quite intuitive a few decades ago.

My Internet handle,

My Internet handle, Spallenzani, is named after Abbe Lazzaro Spallenzani, who played a significant role in the history of microbiology as the first person in history to perform a controlled experiment, at least according to this link.

I was wondering where that was from, although I've usually seen it spelled with an 'a' instead on an 'e'.

Matt, By the way, I should

Matt,

By the way, I should remind people that, hypothetically, global warming seems very intuitive. If you drastically increase the presence of certain chemicals in the atmosphere, it seems intuitive that you would produce some tangible effects on earth. Hence, it's not really akin to a "full moon" type of argument (though oddly enough, there is evidence supporting such a hypothesis.)

But the thing is, good science is skeptical and eschews intuition and 'feel'. Intuition also told people that raindances and virginal sacrifices would work.

As another example, a few years ago, cryotherapy was used to treat stomach ulcers. The sites of ulceration were actually frozen off. It seems very intuitive to think, "Hey, that makes sense. Freezing the tissue will kill the bad stuff." This was a treatment for many years until comparing it to controls revealed no difference in healing rates.

Intuition is ultimately of very little worth in scientific methodology. At best, it can generate hypotheses. But in no way plays a part in proving/disproving anything.

I'm also far from

I'm also far from "knowledgeble" on spelling.

The Montreal Protocol is a

The Montreal Protocol is a precursor agreement to Kyoto, dealing with ozone-depleting chemicals such as CFCs. The phase-out of CFCs has been going swimmingly so far, and signs are that the ozone layer is starting to recover. These facts cause many people to point to Montreal as a successful template for Kyoto. I believe the comparison is flawed, for the following reasons:

1. The damage caused by CFCs (ozone layer depletion) was highly specific and well-established (if perhaps exaggerated), which made it relatively easy to get people to agree on action. This is not true of CO2 and other greenhouse gases, whose effects are highly general, and (at least at this point) difficult to distinguish from normal climate noise.

2. CFCs were used in a relatively small number of (large) industries, mostly as refrigerants and propellants. This made it easy to identify CFC emission sources, and target those industries. CO2, on the other hand, is used in just about every economically productive corner of society. There is no small cluster of "point sources", from which the flow of CO2 can be cut off.

3. There were readily available, economically feasible, easily substitutable replacements for CFCs waiting in the wings--the HFCs. Although slightly lower-performing, they were good enough to allow an easy phase-out of CFCs. No--and I can't emphasize this enough, no--such substitutes are available for hydrocarbons. All such candidates are either politically unacceptable (nuclear fission and soon, I suspect, wind), economically infeasible (nuclear is expensive, and renewables need storage, which adds to their cost and size) or technologically immature (fuel cells running on electrolyzed hydrogen).

There's probably more, but I forget, and I need to work here. In any case, that's Montreal in a nutshell.

Matt, of course

Matt, of course Conservatives are inconsistent in their dealings with the Iraq WMD issue.

I will do more research but here is one link that I found concerning the US role in arming Iraq with WMD during the 80's specifically with mustard gas. http://www.rehberg.net/arming-iraq.html

I don't blame the US for aiding Iraq in the 80's. It is my understanding is was mostly due to counter the religious extremism taking a hold of Iran.

Hey Matt, glad to see you

Hey Matt, glad to see you found the comment threads. :)
it wasn't so hard after all... :)

I'm not sure if your theist argument proves much other than the fact that Karl Popper wouldn't consider theism a science.
Well I think the argument I'm making is weaker than just falsifiability- if I remember correctly that dealt with falsifiable predictions, correct? I'm simply asking for an even that would convince you (or a theist) otherwise, hopefully this side of "god (or whoever) spelling it out in the stars."

Popper thought marxism failed to predict things, hence its failure. I'm just asking marx to say "if x (capitalism producing equality might be one) occurs, marxism is incorrect."

Do you think someone like Kierkegaard cares if his faith is falsifiable?
definitely not. I appreciate the fact, though, that Kierkegaard basically treats the religious stage of existence as insanity, which already takes it out of the running really. I only try this test on theists out to convince me (like the little "Are you going to heaven" booths at every fair in the south.)

Even if we assume that global warming is untestable given our current data, that in itself is not a justification for extremely costly experiments like Kyoto. Perhaps the Montreal Protocol is on a much smaller scale; I've never heard of it before.
I don't mean to suggest that ever batshit scientific hypotheses be tested on a global scale or anything. There is pronounced debate on Global Warming, with many scientists (and several nobel laureates) attesting to human contributions to it. The dangers, as nearly everyone admits (assuming it was true,) are mostly unpredictable but likely very major. Since this is true, I think it's rational to run a sort of global test that could confirm or deny.

As for the intuitiveness of global warming, keep in mind that global cooling was also quite intuitive a few decades ago.
touche'. Apparently popper wouldn't think science was a science.

But the thing is, good science is skeptical and eschews intuition and 'feel'. Intuition also told people that raindances and virginal sacrifices would work.
I wonder about this. Global Warming really stretches science to the limits, for two reasons:
a. the tremendous poverty of data that prevents real climate change knowledge
b. we have a hard enough time explaining how water flows to a sink (no idea,) much less explaining every facet of a complex interrelation of gases in the atmosphere, and how imbalances might affect it.

I think it makes sense to use a broadly intuitive approach, preferring safety over destruction. The idea that "if one greatly shifts the chemical makeup of a thing, features of a thing will likely change" seems like a pretty safe bet to guide me through such scientific torrent. Especially when even the dissenting scientists (GW skeptics) are only clinging to the "we don't really know" argument, rather than the "we know it's not happening" one.

"Hey, that makes sense. Freezing the tissue will kill the bad stuff." This was a treatment for many years until comparing it to controls revealed no difference in healing rates.
please don't understand me as saying that inuition is as good as science or should be used instead. I'm just saying that, given that we're dealing with an issue on the margins of science, between "Humans are causing this problem" and "we don't know if humans are causing this problem, but they sure are releasing shitloads of CO2," intuition might make a good supplement.

I should mention that I intended Montreal as an example of an international, succesful, environmental agreement. I don't disagree that global warming presents new and different problems, I'm only pointing out the we've succeeded in the past at dealing with other similar problems, with tangible results.

if I remember correctly that

if I remember correctly that dealt with falsifiable predictions, correct?

Not quite. Popper was concerned with predictions, but not entirely. More broadly, he was concerned with seperating the disciple of science from other "disciplines" that are pseudo-science. This includes both predictions, methodology, attitude, beliefs, etc.

Popper considered Marxism a pseudo-science not because its predictions turned out to be false (because scientific predictions are falsified all the time), but because Marxists added in ad hoc tweaks to the theory in order to explain why the predictions were not actually falsified. In this way, Marxism, like Freudianism, is able to explain any conceivable situation - it explains too much, and that is its weakness, according to Popper.

There is pronounced debate on Global Warming, with many scientists (and several nobel laureates) attesting to human contributions to it. The dangers, as nearly everyone admits (assuming it was true,) are mostly unpredictable but likely very major. Since this is true, I think it's rational to run a sort of global test that could confirm or deny.

There is more to it than simply demonstrating human influence. We must also determine the degree of human influence and how much it would cost to significantly reduce the human impact on global warming, and whether this cost is worth it. Many economists I have read or spoken to say that under almost every possible scenario, given the uncertainties involved, it is better to spend that money elsewhere, such as on foreign aid, rather than on agreements like Kyoto.

Apparently popper wouldn't think science was a science.

Actually, just the opposite. Popper would have loved predictions like "the world will drammatically cool in ten years," because such predictions are extremely risky, forbid many possibilities, and are held tentatively (or at least should be now that three decades have passed without any cooling). These three factors, among others, make these kinds of predictions extremely falsifiable, which is all Popper was looking for.

I think it makes sense to

I think it makes sense to use a broadly intuitive approach, preferring safety over destruction.

Matt, everyone who is not a sadist prefers safety over destruction, but the "Precautionary Principle" is a shell game. The choice is not between safety and destruction, but between some safety and some destruction associated with choice A and some safety and some destruction associated with choice B. The economic question that must always be asked is, "what are the opportunity costs?" If we spend all of this money on trying to reduce carbon emissions, what is the next best thing that we could have spend this money on instead, and would that alternative save more lives, given our factors of risk?

For example, lets say Kyoto costs $200 billion and has 25% chance of saving 50 million lives. (I just made these numbers up for the sake of the argument.)

Now suppose that direct foreign aid (or some other program if you wish) costs $200 billion and has a 90% chance of saving 20 million lives.

The expected lives saved of the first scenario are 12.5 million compared with 18 million in the second scenario, and both costs the same amount.

Which option should we choose? If our concern is saving the greatest number of lives, given our uncertainty, we should choice the second scenario.

Now it is possible that the uncertainty surrounding global warming will become less so, and the more efficient policy will change accordingly. But it is important to realize that choosing Kyoto is no safer than choosing another alternative, unless we first ignore opportunity costs.

My post dealt with how to

My post dealt with how to make observations about the world, i.e., that anecdotal observations are far inferior to controlled observations.

Popper made contributions to a different part of epistemology: the broader implications what the observations all mean, and how theories - explanations that bring different obersevations together - arise, evolve, change, become obselete, and give rise to better theories.

The scientific method is focused on making observations; the philosophy of science deals with the questions of what the observations mean in a larger sense.

The biggest problem I see with the global warming debate is that the observations themselves are not made very well and with little skepticism.

See

See http://www.gsfc.nasa.gov/topstory/20020926landcover.html for a different causative factor.

Matt, two nitpicks: 1. I

Matt, two nitpicks:

1. I should mention that I intended Montreal as an example of an international, succesful, environmental agreement. I don't disagree that global warming presents new and different problems, I'm only pointing out the we've succeeded in the past at dealing with other similar problems, with tangible results.

I understood what you intended by using Montreal as an example, but I argued that Montreal's success is irrelevant to Kyoto. Any similarities between the two are entirely superficial.

2. ...I think it's rational to run a sort of global test that could confirm or deny.

What did you have in mind here?

Micha, regarding your

Micha, regarding your foriegn aid comparison, I think it's practically inapplicable. The problem is with estimating harms of global heating- there are a few that we can name, but most scientists claim that the effects are largely unpredictable given what we know about homos/heterospheres. If you really look at what most physical science does for envirnomental phenomena, we're not even close to handling root causes yet. Noone knows what causes Rossby waves, for instance, which are responsible for the wind patterns, or what causes El Nino (we know that it starts with the reversal of trade winds but try asking what causes that.)

You're closer to the mark with regard to foriegn aid, the benefits of which could be concievably estimated. Just to play the game though: suppose we could prevent massive glacial melting. It's difficult to estimate how many people could be saved by doing such a thing, but a rough estimate would probably be astronomical, if it included future generations.

Brosef,

I understood what you intended by using Montreal as an example, but I argued that Montreal's success is irrelevant to Kyoto. Any similarities between the two are entirely superficial.
I think your point about HFC's is an important one. Nevertheless, I don't think you've shown the comparisons to be superficial, you just said this one has a few more monkeywrenchs in it.

2. ...I think it's rational to run a sort of global test that could confirm or deny.

What did you have in mind here?
to be honest, I haven't nearly the scientific knowledge to even try and postulate such a thing. I could, perhaps, be wrong but I assume that if x(human produced CO2 in the atmosphere) increases global warming then 2x/3 would see a tangible reduction in such warming.

Matt, I'm not Brosef, but

Matt,

I'm not Brosef, but never mind.

Nevertheless, I don't think you've shown the comparisons to be superficial, you just said this one has a few more monkeywrenchs in it.

I think "monkeywrenches" understates my argument by a rather huge degree. I explained the main reasons for Montreal's success, and showed that none of those reasons apply to Kyoto. Your argument appears to be that agreement and good intentions are both necessary and sufficient conditions for Kyoto's success. But even if everyone agreed, and we had the international will to act, we'd still have the near-insurmountable problems of a) no feasible near-term substitutes, and b) the indispensability of hydrocarbon-based technologies to our economies.

...I assume that if x(human produced CO2 in the atmosphere) increases global warming then 2x/3 would see a tangible reduction in such warming.

But this gets to the heart of Jonathan's argument: there are no control groups. Climate is a highly complex, poorly understood dynamic system. It's very unlikely that we could draw coherent conclusions from such an experiment, due to the impossibility of holding everything else constant. In any case, Kyoto (assuming this is the "experiment" you favour) requires only that the developed (Annex I) nations reduce their CO2 emissions by about 6-7% below 1990 levels, and even that target is likely unachievable. It places no obligations at all on developing (Annex II) nations, who have the fastest growing emissions. Kyoto, even if fully implemented by all Annex I nations, won't go anywhere near 2x/3. I ask again: what kind of experiment did you have in mind?

I'm not Brosef, but never

I'm not Brosef, but never mind.
but you gotta admit that's a pretty funny name. That and brosephine.

Your argument appears to be that agreement and good intentions are both necessary and sufficient conditions for Kyoto's success.
I think that's basically true actually.

But even if everyone agreed, and we had the international will to act, we'd still have the near-insurmountable problems of a) no feasible near-term substitutes, and b) the indispensability of hydrocarbon-based technologies to our economies.
This is a problem of method. Montreal was able to replace the CFC's with something somewhat economically viable- so what? Does that mean the hole in the ozone layer was just tough shit if HFC's didn't exist? Certainly it would've been a good idea to reduce such emissions regardless, am I correct?

But this gets to the heart of Jonathan's argument: there are no control groups. Climate is a highly complex, poorly understood dynamic system.
exactly. If we reduced these emissions and saw improvement, I think it'd be even more obvious that global warming exists. By the way, I feel like a moron writing sentences like "global warming exists" because noone disputes it. The dispute is whether humans are affecting it. I wish there was a good short way to say "human affected global warming."

It's very unlikely that we could draw coherent conclusions from such an experiment, due to the impossibility of holding everything else constant.
Really? The "ozone-layer" skeptics (of which there were plenty let's remember) have certainly been quieted since Montreal. Even though it hardly held everything else constant.

In any case, Kyoto (assuming this is the "experiment" you favour) requires only that the developed (Annex I) nations reduce their CO2 emissions by about 6-7% below 1990 levels, and even that target is likely unachievable.
What does it mean to be "unachievable"? That's nonsense- we could reduce it to 90% if wanted to.

It places no obligations at all on developing (Annex II) nations, who have the fastest growing emissions.
well actually annex II nations are to provide financial support. But yeah, what of it? There is a reason for that, a good one I think.

Kyoto, even if fully implemented by all Annex I nations, won't go anywhere near 2x/3.
exactly. The Kyoto protocol is very weak, but it may well be enough to provide tangible evidence, provided that we cooperated (basically THE neccesary and sufficient condition.)

We have an economic system which, at best, rewards people with "votes" in accordance with the money they bring to bare (how do you spell "bear" in that phrase?) Our grandchildren don't get a vote, naturally, so it's clear that we'll have to enact some legislation.

Micha, regarding your

Micha, regarding your foriegn aid comparison, I think it's practically inapplicable. The problem is with estimating harms of global heating- there are a few that we can name, but most scientists claim that the effects are largely unpredictable given what we know about homos/heterospheres.

Then that is a strong argument against devoting scarce resources towards solving an unpredictable problem - resources that could be used to solve knowable problems. None of this is to say that we should be forcing tax payers to pay for foreign aid, but if we were going to be making decisions collectively (and coercively), it would be wiser to solve problems we know can be solved rather than play craps and hope we win.

If we reduced these

If we reduced these emissions and saw improvement, I think it'd be even more obvious that global warming exists.

But how would we know whether this is causation and not just correlation without a control? Does it make sense to severely curtail the functioning of an advanced industrial economy in order to perform such an experiment that doesn't even have a control? I know you have no problem with this, because you believe in socialism, but why would anyone else agree to this arrangement?

Our grandchildren don't get a vote, naturally, so it's clear that we'll have to enact some legislation.

How is this clear? Even if we adopt a Rawlsian egalitarian outlook, that leads us to the exact opposite conclusion. If you look at the progress of this past century, each successive generation is wealthier than their parents. This indicates that we should be redistributing from the future to ourselves, not the other way around.

mgl: Your argument appears

mgl: Your argument appears to be that agreement and good intentions are both necessary and sufficient conditions for Kyoto's success.
matt:I think that's basically true actually.

And if wishes were horses...

Certainly it would've been a good idea to reduce such emissions regardless, am I correct?

Sure--provided that the benefits of doing so exceeded the costs, and that there were no better competing uses for those resources. Reducing emissions--of any kind--is not a worthy goal in and of itself; it is only worthy to the extent that we realize a benefit from doing so.

If we reduced these emissions and saw improvement, I think it'd be even more obvious that [anthropogenic] global warming exists.

I think your idea of how the climate works is a little simplistic, from what I've read, but I'm not an earth scientist either, so I'll leave it be. The question comes back to Misha's post on opportunity costs, which I note that you haven't addressed. What would we have to give up in order to run this gigantic, uncontrolled "experiment"? What opportunities would have to be foregone? How can we be sure that our grandchildren won't suffer more from the tremendous diversion of resources to your "experiment" than from the putative effects of climate change?

The "ozone-layer" skeptics (of which there were plenty let's remember) have certainly been quieted since Montreal. Even though it hardly held everything else constant.

See my first post on Montreal. The damage done by CFCs was highly specific, and consisted of relatively uncontroversial (and easily simulated) chemical reactions in the planet's upper atmosphere. That's much closer to controlled conditions. The ozone layer is orders of magnitude less complex than the earth's climate.

That's nonsense- we could reduce it to 90% if wanted to.

I think I get it now. If you're stating that something is physically possible, given our current levels of technology, I agree entirely. We could pave the Sahara, build a giant subsea amusement park in the middle of the Pacific Ocean, turn our interstates into flowerbeds, build every person a custom Rolls Royce, cover the Great Plains with 500-metre high wind turbines, and so on. If we wanted to. That's trivial. The interesting question is what we would have to forego in order to achieve these things.

But yeah, what of it? There is a reason for that, a good one I think.

Sure. They're not wealthy enough (yet) to bear the enormous costs of emissions reduction. But that wasn't my point. My point was that, before long, the developing countries will account for the lion's share of CO2 emissions, making Annex I's already-small CO2 reductions look very trifling indeed. Kyoto, at best, will slow by a very minute amount the warming trend; it's not gonna come anywhere close to climate stabilization or reversal. One of the IPCC lead scientists works at the same university as me, and he contends, IIRC, that full ratification of Kyoto would lead to temperatures 0.1?C below the no-Kyoto baseline. In 2100. For this, we propose to conduct a giant uncontrolled, high-risk "experiment"?

Then that is a strong

Then that is a strong argument against devoting scarce resources towards solving an unpredictable problem - resources that could be used to solve knowable problems.
no that's pretty silly. Notice you said "unpredictable" in the first place and "knowable" in the second. Global Warming is very likely "knowable," but when we "know" it it'll be much too late. There are predictions and observations about water shortages and species depletion and such things. Outside of that, and more importantly, we do know a few things that allow us to imagine terrible harms, as well as concrete reasons why we can't predict them (in the words of the honorable Donald Rumsfeld "they are known unknowns.")

We know, for instance, that earth seems to be an incredible balance of forces in nearly perfect position wrt the sun. Not in an "argument by design" sort of way either. While we may have a hard time proving exactly what would happen if we were further from the sun, we can just look at mars and know that if Chocalate is causing the earth to spin out of orbit, we should stop eating chocalate (or "we should devote scarce resources to consumption of chocalate problem" if you grant argument points for opacity.)

But how would we know whether this is causation and not just correlation without a control?
well being the rational scientists that we are, perhaps we should just reproduce the entire earth but not cease CO2 emissions to get a nice control. A believe it's you who's so fond of the phrase "don't let the perfect be the enemy of the good." We're interested, fundamentally, in not having the earth increase in temperature to any significant degree- not in testing the validity of global warming theories. If we stop that increase, then break out the milk and cookies.

If you look at the progress of this past century, each successive generation is wealthier than their parents. This indicates that we should be redistributing from the future to ourselves, not the other way around.
I assume you mean this as a joke. Not withstanding the impracticality of such a thing, or the fact that you've adopted a rawlsian egalitarian outlook, which is clearly absurd for you, we've still got the argument that we might not have a 3rd generation from now because of a few things. If I were in the original position, I doubt I'd choose to live in Waterworld (supposing that the original position included time machines.)

mgl, I'll hit your post tommorrow or at least by Monday. I'm checking out a few articles from the library, because it seems true that scientists actually do agree that humans are causing it anyway, possibly making this hypothetical argument irrelevant.

ps- "A study by scientists

ps-
"A study by scientists at the World Health Organisation found that 160,000 people die every year from side-effects of global warming, ranging from malaria to malnutrition, and the numbers could almost double by 2020." - Reuters, London

got a little more

got a little more time:

Sure--provided that the benefits of doing so exceeded the costs, and that there were no better competing uses for those resources. Reducing emissions--of any kind--is not a worthy goal in and of itself; it is only worthy to the extent that we realize a benefit from doing so.
CBA can only take you so far you know. Of course there's nothing axiomatic about reducing emmisions- that's hardly the area of contention. This is more a "precautionary principle" argument- an argument that's conservative (in the actual sense of the word.) We're not in a position to really quantify all the benefits (though certainly we can know some of them, ala my last post,) so CBA just doesn't apply. That doesn't mean we can't make a decision though, and it would be interesting to spell out the quantifiable harms in terms of CBA, too.

I think your idea of how the climate works is a little simplistic, from what I've read, but I'm not an earth scientist either, so I'll leave it be. The question comes back to Misha's post on opportunity costs, which I note that you haven't addressed.
perhaps I got it with the last response? If not, just remind me.

What opportunities would have to be foregone? How can we be sure that our grandchildren won't suffer more from the tremendous diversion of resources to your "experiment" than from the putative effects of climate change?
Matt: "There's a monster coming! Let's get an army together and fight it off!"

mgl: "what about the people who will fight in the army? In the time they would've had they could've concocted a miracle drug allowing us to live till we're 150, and that'd be more important than some stupid monster."

Micha: "Yeah, life expectancy is a knowable problem, whereas the deaths caused by a monster are largely unpredictable, so we can ignore it."

Discovery Magazine (from this month I think): "people born today have a good chance of living to 150 years, a tidbit which inspired matt's use of '150' years above."

My point was that, before long, the developing countries will account for the lion's share of CO2 emissions, making Annex I's already-small CO2 reductions look very trifling indeed.
that's your point? I actually have a few problems with the "annex" situation- China being included in Annex II is a big one- but that's not a good argument at all. Once the developing countries were producing such things, we could update the annex status (Montreal has been updated like 6 times I think.) That's like saying that because kids will become adults we don't need to legally distinguish between the two. Or actually (if I follow your argument) it's like saying that because kids will become adults, we don't need a court system at all!

One of the IPCC lead scientists works at the same university as me, and he contends, IIRC, that full ratification of Kyoto would lead to temperatures 0.1?C below the no-Kyoto baseline. In 2100. For this, we propose to conduct a giant uncontrolled, high-risk "experiment"?
what does IIRC mean? But I do agree that Kyoto is a pretty weak treaty, and it's pretty pathetic that we can't do even that. Kyoto would just get the ball rolling- then it could be strengthened.

Outside of that, and more

Outside of that, and more importantly, we do know a few things that allow us to imagine terrible harms, as well as concrete reasons why we can't predict them (in the words of the honorable Donald Rumsfeld "they are known unknowns.")

You can't have it both ways. Either the costs of global warming are inherently unpredictable, in which case any decision is simply guess work with no rational calculation, or it is somewhat predictable and we can attach expected costs and benefits given various interests rates and risks. The second option lends itself to cost benefit analysis. The first lends itself to tea-leaf reading. If global warming is inherently unpredictable, then scarce resources will not be diverted away from predictable uses, unless we decide to make economic decisions completely at random, which we do not do, at least with tax dollars.

This is false, of course, because people do make predictions about the various costs and benefits associated with global warming, and these figures can them be used in comparison to the next best alternative. In which case, we can reject your "broadly intuitive approach, preferring safety over destruction" because all the alternatives have a risk of safety and a risk of destruction. The question is which alternatives are better and which are worse.

We're interested, fundamentally, in not having the earth increase in temperature to any significant degree- not in testing the validity of global warming theories. If we stop that increase, then break out the milk and cookies.

But we are also interested in various other things we can do with those resources instead of spending them on Kyoto. The question is which alternative is a better investment. Not having a control makes investing in mass world-wide experiments a sketchy deal.

"If you look at the progress of this past century, each successive generation is wealthier than their parents. This indicates that we should be redistributing from the future to ourselves, not the other way around."

I assume you mean this as a joke. Not withstanding the impracticality of such a thing, or the fact that you've adopted a rawlsian egalitarian outlook, which is clearly absurd for you, we've still got the argument that we might not have a 3rd generation from now because of a few things.

Not a joke at all. This is taken directly from environmental economics. I don't have to adopt a Rawlsian outlook to use it as part of my argument to show that even according to my opponents' premises, Kyoto is a bad idea. And it is not impractical to redistribute wealth from the future to the present; all we have to do is save less resources for our children by using them now.

But I do agree that Kyoto is a pretty weak treaty, and it's pretty pathetic that we can't do even that. Kyoto would just get the ball rolling- then it could be strengthened.

Kyoto is a pretty weak treaty precisely because everyone agrees that the costs outweigh the benefits. Instead, you are arguing that Kyoto is a kind of beneficial slippery slope, not good in and of itself, but good because it will inevitably lead to better treaties in the future. This is silly. The reason why people are hesistant to sign Kyoto in the first place is because it is a bad deal; if we can come up with a better deal, why not suggest that instead?

What I want to know is why

What I want to know is why there is no mention of Bayes theorem which can handle this situation quite nicely?

Ultimately we are intersted in the question:

Are we in state S1 or S2. Thus, the thing we are interested in is:

Prob(S1|E), and
Prob(S2|E).

The notion of control in experiments is somewhat overblown, IMO. Can you control each iteration of a experiment perfectly? No. Yes, you can remove many confounding factors, but even small ones can have problems.

Also, the idea that you'd look at a single data point is specious reasoning. Sure, a hot summer in Europe might be more likely with man-made global warming. But there is also little warming trend in North America. There is no warming trend in the lower troposphere. The land based record does show warming, but it too has its problems.

Also, lets note that the two hypotheses are not mutually exclusive. S2 also covers the state where you have global warming due to non-man-made reasons. When you do this kind of analysis you have to look at all the data and revise each time you get new data.

IIRC = IF I Recall Correctly

what does IIRC mean? But I do agree that Kyoto is a pretty weak treaty, and it's pretty pathetic that we can't do even that. Kyoto would just get the ball rolling- then it could be strengthened.

Yeah, yeah. Sure, lets really tank the world economy.

Link