Theme Thursday: Climate Change   19 comments

bad-300-250

I’ve been interested in science, mathematics, and philosophy for a long time now.  At various points in my life, I’ve been a practicing mathematician and an armchair philosopher, and I’m currently engaged in scientific research.  I’ve worked for an educational institution specializing in hard core research for almost a decade now.

I’ve known scientists for a long time.  I’ve seen grumpy scientists and happy scientists, conservative scientists and liberal ones (somewhat more of the latter than the former).  I’ve seen scientists have arguments with each other just this short of breaking out firearms and elevating to mayhem, and those same scientists publishing work together after they discovered that they were both wrong about their relative positions and finding the results of their argument to actually be interesting.

I’ve seen lazy scientists, venal scientists, greedy scientists, and scientists with such an odious personal manner that you’d be hard pressed to share a meal with them.  I’ve seen lots of towering egos, and almost as many deflated ones.  Quite a few reconstructed ones in the mix as well.  [edited to add: lest this sound like all scientists are stereotypical difficult personalities, I’ve also known plenty of nice, friendly, and outgoing scientists.]

I have never seen a scientist actually engage in unethical behavior in their own field.  I know it happens, there are plenty of examples… but on the whole the act of deliberately falsifying results and misrepresenting reality is as uncommon among scientists as running away from a burning building is among firefighters.  Most of them can’t even conceive of the idea; it’s almost incomprehensible.

This is not because scientists are a uniformly ethical crowd (although anyone who will spend their entire life on research that they largely don’t profit from in proportion to the impact of their work is probably going to have at least a high baseline sense of ethics).  It’s because scientists know that reality wins.  If they publish bogus results, sooner or later someone will try to replicate their results, or find some other results that contradict the bogus results.

Scientific argument isn’t like political argument.  The scientific method isn’t even like mathematics.  In mathematics, you declare your axioms and prove interesting stuff follows logically as a result.  In science, you observe reality, make notes, and draw conclusions.  You can have all the nice, logical, consistent theories you want… but when you put a paper up for peer review, or attend a conference, or try to discuss your work with another scientist they’re only interested in the theory in passing.  Reality wins.

What they want to see is the evidence.  Not “beyond reasonable doubt” evidence, but “towering monolithic gargantuan piles of evidence”.  If you don’t have it, you can get eaten alive (at least, metaphorically speaking)… if you don’t have it, you’ve got what is commonly referred to as, “An interesting little theory”, a phrase that itself carries a depth of meaning that isn’t parsed well by people who aren’t learned in the particular field.  The difference between a cap-“T” theory and a little-“t” theory is the difference between the Nobel Prize (not the Peace prize, which has no measurable standard, but the other prizes, that you can only get if you’re the freaking grandmaster ninjitsu tenth-level jedi master gun-kata guru 1,000 lb gorilla of a scientist)… and not getting the invite to go out bar hopping after the keynote speech.

Even if you *do* have a huge pile of evidence, you can get eaten alive if your resulting theory directly clashes with existing theory.  This isn’t because science is hidebound or dogmatic; it’s because scientific theories are based on lots and lots of observations, and if you come up with a new theory that challenges existing theory, you’ve got a pretty high bar to climb over.  People point to all of the major turning points in science as if those moments represented some sort of failure… “See!  They used to think exactly the opposite of what they think now!  They can never make up their minds!”

What those naysayers don’t realize is that “never making up your mind” is a central tenet of being a scientist.  You take some things for granted because nobody has time to learn everything and someone else is better versed than you are, but if someone shows that what you took for granted is wrong, you change your mind.  If you don’t, you don’t get published (at least, not for very long), and that’s the long slow death of the scientist.  Tenure doesn’t mean much if you can’t get a grant.

Reality wins.

Of course, scientific discourse isn’t political discourse.  Scientific discourse isn’t legal discourse.  There’s plenty of studies that show cigarettes cause cancer; it still took decades of fighting misinformation before anyone who worked in the tobacco industry would admit that reality really was described best by the theory that there was a causal link between tobacco and cancer… and it wasn’t piles and piles of scientific studies that convinced anybody, it was a legal and political battle.

The “Climate Change Debate” is just like that.  There isn’t a climate change “debate” among climatologists.  There isn’t even really a climate change debate among scientists in general (a couple of outliers, none of whom study climate science, does not a debate make), nor is there serious belief in a  “non-anthropocentric cause”.  Not because scientists are out to get rid of technology (I have yet to meet one that didn’t love his or her computer).  Not because they’re out to halt progress (you pretty much need to be on board with the *idea* of progress to become a scientist in the first place).

It’s because the evidence for other theories isn’t there.  There are several major research journals in climate science, and (unsuprisingly) the Caltech library subscribes to electronic versions of them all.  When someone pointed me at Senator Inhofe’s web site, claiming that there was peer-reviewed science that refuted global warming, I went looking for it.  I didn’t find it.  Irritated that I had no actual citations to start with (something I find to be a general media failure, so I could hardly assume that the lack of such was immediate evidence of nefariousness)  I looked again.  I still couldn’t find it.

So I went out on the general Intranet and tried to find any sort of reference to the actual journals these papers were published in, or what their titles were.  I didn’t find that, either, but I did find a whole pile of blog posts by scientists on Inhofe and dismantling the claim into teeny, tiny little shreds.

If you believe that global warming is bunk, you’re very, very likely to be very, very, horribly wrong.  Not guaranteed wrong, of course.  Again, science is not mathematics.  We can’t say that anything is definitively true, because we don’t know for certain what all the axioms of the Universe are.

So what?

We also can’t say that it’s definitely true that if you stick a loaded gun against the side of your noggin and pull the trigger that you’re going to die.  The gun might not fire.  The bullet might be a dud.  The gun could explode in your hand, and just cause a terrible injury, or the bullet might bounce off your skull or by some random roll of the dice not hit anything critical on its way through your grey matter.  Maybe we’re all plugged into the Matrix and the truth is, there is no gun.

I wouldn’t bet on it.  It astonishes me that so many people are not only willing to bet on it, they’re eager to do so.

Advertisements

Posted October 14, 2009 by padraic2112 in science, Theme Thursday

19 responses to “Theme Thursday: Climate Change

Subscribe to comments with RSS.

  1. mmm…the gun analogy makes a lot of sense. reality wins. you choose to believe th facts or not. the problem is your decision does not just affect your own mortality.

  2. very interesting…I am not a Scientist…I also don’t think I really “know” anything…but I think I have a pretty good idea that something is askew with the whole ball of wax. Thanks for the info!

  3. I don;t think anyone debunks that the earth is warming but if its really from man is hard to prove. Perhaps its another cycle? I don’t know. I do try my best to take car of what I know makes a difference–recycle to reduce land fills expanding and reduce water intake in dry areas.

    • “I don’t think anyone debunks that the earth is warming but if its really from man is hard to prove.”

      You’re making an irrelevant statement. You’re also looking at it backwards.

      Science (at least, non-basic science) is not about “proof” in the binary logic sense. Like I said, I cannot say definitively that if I point a loaded gun at you and pull the trigger that you’re going to die, but the probability is so high that nobody would think twice about convicting me of attempted (or actual) murder.

      There is a huge amount of research on the topic, twenty plus years of studies in multiple fields (climatology, zoology, biology, chemistry, atmospheric science, you name it) that adds weight to the theory. There is no credible falsification data at this point; nothing substantial to provide an alternate theory.

      The affect of carbon dioxide, methane, and other greenhouse gases in an atmospheric medium is basic chemistry, you can perform the experiments using high school chemistry equipment. The rate of addition of these gases to the atmosphere due to human activity is on a massive scale, and has been measured for years.

      The burden of proof is actually on the *negative* side.

      We know these gases make the atmosphere warmer, we can prove that experimentally.
      We know we’re adding these gases to the atmosphere at an incredible rate, we can show that with basic data collection.

      If you want to debate AGW, you’re going to have to come in with a pretty big chunk of evidence… not just providing an alternate cause for global warming (so far, nobody’s even successfully done that much), but also provide a reasonable hypothesis for why human activity creating these gases *won’t* warm the atmosphere.

      What is it about the atmospheric processes of the Earth that might “naturally” reduce the basic chemical properties of CO2 and methane to warm the atmosphere? (answer: nobody has even proposed a credible idea).

  4. Interesting indeed! Thanks for that.

  5. Absolutely great post! I’ve spent decades talking to brick walls who just don’t get the scientific method, and insist on treating all Theory as theories. It can get amusing at times, though, like reading some of the weird contortions the climate change deniers twist themselves into to get around the undeniable evidence. Quite a few candidates for the ranks of the Chinese Circus and the Cirque de Soleil there!

    One of the greatest quotes on the spirit of science comes from an episode of Stargate SG-1, when an off-world scientist says to SG-1 member Teal’c: “Teal’c, I’m a scientist. When I find evidence that my theories are wrong, it’s as exciting as if they were correct.”

  6. Great post and comments. Thanks.

  7. Great GREAT post. You think, you think, you think. No way, not likely, likely, very likely, sure thing. Every ting falls on the line. -J

  8. Pat-

    One of the problems with making science ‘definitive’ in the sense that you can reasonably establish the likely ‘truth’ of a phenomenon is that there is still a woeful lack of attention paid to solid experimental design coupled with overuse of ‘cookie cutter’ statistical analysis in Systat, SPSS, or any of the other commercially available stats packages out there. I took an experimental design and advanced stats class in graduate school and one of the very interesting trends we looked at was the number of papers published (in big name peer reviewed journals, no less) with weak stats and poor design. As it’s fairly typical to make statements such as “P=0.05 or less, so therefore this is significant/true/etc.”, weak stats can significantly change the results of the analysis.

    Couple that with the fact that it’s not possible to approach the problem in a truly experimental fashion (as we can’t isolate variables, establish controls, etc) and it becomes even harder to conclude anything meaningful. Sure, we can create models (which may or may not be useful based on our incomplete understanding of all variables and how they interact), but they can sometimes be unconvincing, as it is too easy to tickle the parameters of the model to produce a particular outcome. So we’re left, really, with a questionable data set from weather stations around the world with samples taken by different researchers potentially introducing their own sample bias. Sure, we can look at ice cores, but we can’t currently derive temperatures over time from them.

    But as you suggest, the precautionary principle is worthy of consideration here. Global climate change (if it is happening) might potentially force drastic changes in human lifestyles at a tremendous cost in life and resources. At worst, investing in climate change studies and technology will give us more information about being good global stewards. At best, it might save our collective bacon.

    But I don’t expect that the naysayers will go away any time soon. Consider the anti-immunization pundits that are still around.

    • > I took an experimental design and advanced stats class
      > in graduate school and one of the very interesting
      > trends we looked at was the number of papers published

      Certainly, one of the problems with a good number of the existing scientific fields is a heavy reliance on statistics without a correspondingly high understanding of what constitutes good statistics. Note, though, while this certainly does inject a number of substandard papers into the publishing cycle, it doesn’t necessarily overly affect the body of knowledge as driven by the overall community. I’ve seen the same sort of issue in IS papers… but the people who are really good IS researchers don’t cite conclusions from weak papers. Good results build on themselves. It’s one thing to say, “There are unconvincing papers published in peer-reviewed journals”, it’s another to say, “Those unconvincing papers represent a significant bias, as they unduly influence the publication of other papers”.

      Log into ISI Web of Science and you’ll see a bazillion papers on “community response to large scale disasters”… but most of the garbage papers aren’t cited by anybody else, and most of the good ones are citing by lots of people. Measuring the significance of research isn’t a boolean function. Peer reviewed journals are only one step in the construction of scientific knowledge (a pretty big step, sure).

      This in and of itself also begs the question… if it’s not impossible to publish weak results, how come we don’t see weak results published that disconfirm AGW?

      > Sure, we can create models (which may or may not be
      > useful based on our incomplete understanding of all
      > variables and how they interact), but they can
      > sometimes be unconvincing, as it is too easy to tickle
      > the parameters of the model to produce a particular
      > outcome.

      This I’ll disagree with. It’s important to be skeptical of models, sure, because the models themselves depend heavily on initial conditions. It’s also difficult to estimate the efficacy of models that predict major phase shifts. However, a good model does create a falsification methodology, because the outcomes of the models can be compared to other data sets (not just the primary outcome, but secondary outcomes). There’s been lots of limited climate models, but the current “big players” have a pretty well developed methodology. Running those models produces expected temperature gains, and expected carbon deposits. We find the carbon deposits in the ice cores. *Not* finding the carbon deposits in the ice cores would be a falsification of the model.

      Judging the efficacy of the climate model itself requires a lot of input from people who study climate models. Most of these guys have thrown weight behind a very limited subset of climate models. Sure, this means that there’s lots of bad models, but that doesn’t mean that the models that *are* being used as keystone reference models don’t work 🙂

      Direct observation is definitely warranted, at this point… and it’s happening (albeit not by the U.S.) – the Gosat project is up and running (http://www.gosat.nies.go.jp/index_e.html).

      • You have to be really careful not to confuse ‘good’ with ‘popular’ or ‘easily acceptable by the community’ or ‘published by a recognized authority in the field’- scientists are not without bias. We found plenty of ‘foundational’ papers in the biological sciences that had conclusions based on methods or analysis that left plenty of room for doubt. Consider that journals themselves represent a significant source of bias, as a select few editors decide what is and is not worthy of publication. One of the big tenets of the scientific method is that one should question everything, that there is no such thing as an unassailable construct. So the moment you, as a practitioner of the scientific method, start saying “this idea is BS” you are moving towards ‘science as cult’.

        Modeling is another example of this problem. You can create intricate models that incorporate as much data as you can possibly imagine and you’ll always be shy of the mark. Just look at weather forecasting; there’s an entire scientific discipline devoted to the study of weather, and our models are significantly inaccurate even 24 hours out. Models have the most value in helping us understand how phenomenon work and not as predictors- unless your phenomenon is simple and fairly mathematically precise. Tide tables would be a good example, but even they are inaccurate now and then.

        Again, I contend that the climate change debate can’t be about the data or the models, because neither one is going to stand up to reasonable and rigorous scientific deconstruction. If they could then we wouldn’t be having this conversation. I maintain that the best arguments in favor of climate change mitigation are 1) the precautionary principle, and 2) the value in practicing good stewardship as a holistic approach to species- and biome- level resource management.

  9. > You have to be really careful not to confuse ‘good’ with ‘popular’
    > or ‘easily acceptable by the community’ or ‘published by a
    > recognized authority in the field’- scientists are not without bias.

    Sure. That’s basic psychology; people are likely going to believe more in the evidence that supports beliefs they already hold. This happens in science, too… but it is to a decidedly lesser extent in the medium and long run. In the short run, people will disbelieve and discard contrary evidence. In the medium to long run, contrary evidence doesn’t go away, and eventually it builds due to repeatable results.

    > We found plenty of ‘foundational’ papers in the biological sciences
    > that had conclusions based on methods or analysis that left plenty
    > of room for doubt.

    Well, you were the biologist, so I’ll take your word for that. I’m not so sure that “plenty of room for doubt” necessarily means the same thing in all fields, though.

    > Consider that journals themselves represent a significant source of
    > bias, as a select few editors decide what is and is not worthy of
    > publication.

    Careful, this is a journal-specific problem, so “significant” means different things for different journals (and different fields). Some journals are better at this than others. You can’t color the entire scientific publishing field with one brush here. Editors, as a class, aren’t the Inner Sekret Cabal Keyholders. And again, this is a shorter term problem. Editors die, get replaced, go on sabbatical, quit, are fired, etc. At worst, an editorial bias is going to inject a frequency bias towards a particular theory. Hell, in my field editorial bias is almost a non-issue, because the good stuff is in the conference papers, as technology moves so fast the likelihood of a paper being significant through a conference review/presentation/edit/journal submission/review/edit/publish cycle is pretty low 🙂

    > One of the big tenets of the scientific method is that one should
    > question everything, that there is no such thing as an unassailable
    > construct.

    I hear this alot, and many people think it means something other than what it actually means. Not saying that you do, but for other readers of the blog…

    Yes, you shouldn’t regard anything as an unassailable construct. But at the same time, “question everything” is a reducto in the opposite direction. When you’re studying empirical behavior, you obviously can’t account for all factors, but if you develop a repeatable framework describing a wide class of behaviors, there’s no reason to “question everything” all the time, in fact, you’ll wind up acting like Hume or Popper and claiming that there’s no basis for induction; you can only ever say anything about something that’s already happened, and you can never claim predictive ability. This is just stupid; there’s lots of ways to confirm that you *do* have predictive ability for certain types of phenomena.

    The best example of this is Newtonian mechanics. We know Newtonian mechanics is actually wrong, as it breaks down both at the macro and micro level (relativity and quantum mechanics, respectively). However, for most phenomena (which are well within those boundary conditions), it’s perfectly acceptable to assume that Newtonian mechanics is actually true, because it is… within the boundaries of the class of phenomena that it actually describes.

    > Just look at weather forecasting; there’s an entire scientific
    > discipline devoted to the study of weather, and our models are
    > significantly inaccurate even 24 hours out.

    This is actually a bad example, since you’re talking about a boundary condition behavior. Localized weather patterns are hugely susceptible to complex local conditions, which is why a tornado warning can be inaccurate in a given small geographical region. Given the type of behavior you’re modeling, you’re actually getting pretty accurate predictions. Climate is a whole ‘nuther story, because most of the complex local conditions become less statistically significant as you increase the time differential.

    Put another way, predicting that a tornado is going to occur within a particular county on a given day is hard. Predicting that a number of tornadoes are going to occur in a state in a given decade is quite a bit easier.

    > Again, I contend that the climate change debate can’t be
    > about the data or the models, because neither one is going
    > to stand up to reasonable and rigorous scientific
    > deconstruction.

    Heh. Here I’d say that you don’t have the credibility to make such a claim for the particular field. You don’t know enough about the model design, the methodology, or the mathematical competency of people who design and run the models in question. Now, I’m not a climate scientist, but I’ve seen mathematical critiques of a number of different climate models that are used as the baseline models, and they’ve been given good marks. Since I’m not a climate scientist (and since the mathematical critiques can only talk about the math, not the inputs), I can’t make a contention that the inputs are reasonable for the models, but when a few thousand climate scientists all say they’re pretty good and nobody in the field says they aren’t, I’m okay with assuming that they’re not all total boneheads. That to me is a reasonable and rigorous deconstruction.

    Now, in your own field, mathematical models are most frequently used to study stability of fish populations, right? And what were the good models saying, for years…? They were saying that a good number of fisheries were close to collapse, which has been borne out by the fact that those fisheries have actually collapsed. Now, they may not have been the popular models then, but I imagine they’re the popular models now, since they have confirmation.

    A lot of the climate science modeling *does* have verification (just look at the Wilkins Ice Shelf collapse, or the glacial retreat, or the melt rate in Greenland, etc.)

  10. You may very well be right, from a public policy standpoint focusing on something other than the science may be the most effective route, but I’m ambivalent towards this approach. In the particular case of smoking, that’s actually what *worked*, as it just took 30+ years of litigation to change public perception, and not a couple years of scientists saying, “Hey, dummies, sucking ash and smoke into your lungs is stupid”.

    But for all its short term biases and occasional drifts down the wrong explanatory paths, science works. It’s the reason we have the technology we have. People need to understand why it works, how it works, and how it corrects itself before they’ll trust it. Punting away from scientific questions because explaining the details to laymen is hard gets you into the situation we’re in now regarding vaccinations: people don’t trust them, because they’ve been told repeatedly that they ought not, and scientists have sat in their little science silo and said, “these people are ridiculous and we don’t need to pay any attention to them because we know the science”.

  11. Oh one more bit on this:

    > We found plenty of ‘foundational’ papers in the biological sciences
    > that had conclusions based on methods or analysis that left plenty
    > of room for doubt.

    Remember that most theories are laughably incomplete when they’re first published, they don’t generally spring fully-formed from the head of their source (at least, not since the 1800s). So foundational publications don’t just become foundational publications because they’re completely iron-clad, they become foundational papers *over time* as additional results support the original claims.

    Somebody can make a prediction based upon a theory framework and a substandard model, but if the prediction comes true, someone else cites the study, builds a better model to extend the predictive ability based upon the same theory framework, and tests their own results. If *that* prediction comes true, the process repeats itself.

  12. Trying to follow an Andy-Pat exchange always makes me feel as if I’m still up there swinging around in the tree, while they’re down there tool-using…

  13. Um, not the tool that might come to Hammer’s mind, though…

  14. I watched an Inconvenient Truth, then the Great Global Warming Scandal (oh yes, I know where to get my peer reviewed scientific knowledge!) and both had scientific umph behind them . . Climate Change is perpetual. It’s happened before and will again. My son did a paper during his Horticultural Degree and carbon evidence in ice cores shows this to be true. Bottom line is hey, clearly things are awry. Over consumption is rife, the planet’s looking a little poorly so shouldn’t we err on the side of caution and assume a fully functioning loaded gun pressed against our temples? Interesting post, love the argument between you and Andy although now as usual, my head hurts and I needed toast with Vegemite. I tend to think that it’s best to assume the worst and fight against it. Hey, Cancer might kill me . . it might not but I’m eating anti-oxidents like they’re going out of style – just in case.

  15. > Climate Change is perpetual. It’s happened before and will again.

    Oh, sure. In fact, one can certainly argue that learning how to terraform our own planet is a damn good idea, seeing as how there have been several major extinctions in the Earth’s history. The whole “Gaea Earth Mother” schtick is crazy; the Earth doesn’t give two cents for any of the lifeforms that have been generated over the eons, it’s not going to start with us.

    Fun reading: Phil Plait’s “Death From The Skies”.

  16. Right now, it’s snowing outside and has been all day. Even for Massachusetts this is wierd. Global warming , climate change, whatever you call it, we have a lot to do with it. Hope more people wake up before it’s too late.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: