Andy May on the Philosophy of (Climate) Science
Andy May recently wrote an opinion piece in the Washington Examiner[1] that I think warrants a rebuttal. If you haven't heard of him, Mr. May is a petroleum geologist who somewhat inflates his credentials when sharing his opinions to obscure his conflict of interest with climate science. He claims that he is a "petrophysicist" (he's a petroleum geologist with a BS in Geology), a "paleoclimate expert" (which he demonstrably isn't, see below), and a member of the CO2 Coalition (which he is). But he's popular in the contrarian blogosphere, so I think there are some things we can learn from interacting with his claims here that can yield some helpful insights into climate science and the scientific method.
In this opinion piece, Andy May is claiming that Karl Popper's philosophy of science should cause us to see that climate change is unfalsifiable and therefore pseudoscience. Well, I think that's what he's trying to say but he kind of blunders it with a misunderstanding of Popper, climate change, and the scientific method. Let's start with this:
It should be obvious, I think, that you can't call climate science pseudoscience because of what politicians and the media say about climate science. If that were true, then quantum mechanics would be pseudoscience. Any evidence supporting this claim should come from the scientific literature, and in particular, peer-reviewed studies. Does the scientific literature present "tornados, hurricanes, mid-winter thaws, 100-degree days, cold snaps, droughts, and flooding creeks" as untestable (or unfalsifiable "evidence of human-caused climate change" or "proof that emissions of carbon dioxide are leading to an overheated planet?" Not in my reading of the literature, and May fails to provide a single example of this happening. Are we to just accept what he says here on his authority as a petrophysicist, paleoclimate expert and member of the CO2 Coalition? I hope not.
What does the scientific literature actually say about extreme weather events and climate change? A fair summary would be that as our emissions increase greenhouse gas concentrations in the atmosphere, we should see its effects on the climate system, including 1) an increase in global mean surface temperature and 2) the warming and acidifying of the oceans. These should in turn affect other aspects of the climate system, including 1) melting of glaciers and the shrinking of sea ice, 2) accelerating sea level rise from thermal expansion of the oceans and the influx of glacial ice, 3) increased uptake of CO2 by plants (greening), and 4) changes in weather patterns, especially extreme weather. Scientists also, through models and empirical methods, explore in what ways these extreme weather patterns will be affected by AGW. These include an increase in extreme heat, increased drought in areas prone to aridity (like in the American southwest) and increased flooding in areas prone to flooding. I could of course support each of these with evidence from the literature, but I think most people even moderately familiar with climate science would agree that the above is common knowledge, and I've interacted with the scientific literature here in other posts.
More importantly, every single one of these are testable theoretical predictions of climate science that demonstrate that climate science is indeed falsifiable. In the absence of stronger forcings pushing climate towards cooling, an increase greenhouse gases pushes climate towards warming. Is that observed? Yes. It should also warm the oceans. Is the ocean heat content increasing? Yes. It should also acidify the oceans. Is the average pH of the oceans decreasing? Yes. It should cause accelerating sea level rise. Is that happening? Yes. It should cause some "greening" due to the uptake of CO2 from plants. Has that happened? Yes. And it should cause changes in weather patterns, and in particular, extreme weather. Are we experiencing these changes? Absolutely. Extreme heat is far more common now than it was in the 1950s. Arid areas are experiencing more droughts and longer droughts, causing water shortages and increases in wildfires in these areas, while other areas are experiencing more flooding. More examples can be given here, but suffice it to say that, even if you disagree with me that these theoretical predictions have been confirmed with empirical data, it's absolutely clear that climate science contains theoretical predictions that can be tested against experience. Climate change is falsifiable.
What I think May may be getting at, if I give him the most charitable reading possible, is that it may appear that when unusual weather happens, that unusual weather is automatically blamed on climate change. If this is what he means, it's not "climate science" as a whole that is pseudoscience, but the aspect of climate science that attributes causation to weather events. He seems to think, whether you experience a heat wave or a cold snap, the event will be blamed on climate change. Whether you experience a drought or a flood, the event will be blamed on climate change. If every effect is automatically climate change, then this aspect of climate change can't be testable, and therefore it can't be scientific. But even this is not a fair assessment of the scientific literature (though some news stories do this). The scientific literature does not simply say we had bad weather, therefore AGW. Rather, the literature looks at how AGW theoretically should impact various regions, and that allows for testable theoretical predictions. So in the American southwest, an increase in temperature should deplete soil moisture by increasing VPD, which in turn should make droughts more frequent and severe. This is a testable prediction. And should it not be observed, theory can and should be revised to account for the reality. That's climate science functioning as a science. May provided literally no evidence of climate science itself operating as a pseudoscience.
What May does instead is wax eloquent about Karl Popper, though he doesn't get Popper quite right. He correctly points out that one characteristic that should be built into all scientific theories and hypothesis is testability or falsifiability. To investigate a theory of hypothesis, there needs to be a way to test it against experiment. So scientists develop theoretical predictions that can be tested against empirical evidence. Should it fail a properly conducted test, then the theory or hypothesis is wrong, and it must be discarded or modified in light of the new evidence. Any theory or hypothesis that can't be tested in this way is not properly scientific, which is why climate scientists go to such great lengths to test the theoretical predictions of climate science against empirical data.
But from here, May goes a bit wonky: "Popper draws a bright line between science and pseudoscience. The line is falsification. Pseudoscientific ideas cannot be falsified." The last sentence is absolutely wrong. There are many pseudoscientific ideas that can be falsified. It's clearly a pseudoscientific idea to say that the Earth is flat, but that is absolutely a testable claim - it is falsifiable, and it has in fact been falsified. Many pseudoscientific ideas are falsifiable. While it's correct to say that unfalsifiable claims don't qualify as scientific, many claims that qualify as being falsifiable are still pseudoscientific. Among falsifiable claims perhaps a good way of discerning pseudoscientific falsifiable claims would be 1) holding onto claims after they have been falsified or 2) rejecting claims without compelling evidence that have been confirmed by overwhelming evidence. As an example of the former, the cosmic ray hypothesis began as a legitimate scientific hypothesis, and it has been rigorously tested. It failed in part because trends in cosmic rays have been running counter to observed warming. There was absolutely nothing wrong with the scientists that proposed the hypothesis - we can learn a lot from proposing these hypotheses, even when they fail. What makes it currently pseudoscience is that those who hold to this view are holding onto a hypothesis that has already been falsified. Likewise, an example of the latter would be the rejection of the falsifiable claim increasing atmospheric CO2 pushes the climate system towards warming. This claim has been overwhelmingly tested against empirical evidence, and it has passed. It is not proven - there is a non-zero chance it could fail some test in the future - but the evidence for it is so strong it's pseudoscientific to reject it without also supplying compelling evidence overturning the evidence we have.
Can we find pseudoscience in larger discussion of climate science? That is, is there anything that qualifies as being comparable a Marxist searching for evidence confirming his theory in the newspaper and ignoring counter evidence? Absolutely. One of the more popular examples of pseudoscience is the claim that warm climates always correlate with thriving empires and cultures, while cold climates always correlate with societal collapse. This is a pseudoscience claim that fits exactly the example of a Marxist searching for confirming evidence for Marxism in a news paper. All you need to do is look for when temperatures are warm and find thriving empires and cultures that line up with those warm periods. Then you look for times when temperatures are cold and look for examples of societal collapse. Then you label the a graph showing these cherry picked events correlating with warm and cold periods, and there you go - pseudoscience at its finest.
It turns out, Mr. May has been doing this for quite some time. Back in 2016, he published a graph on his blog[2] showing "evidence" that warm periods are good for culture and cold periods are bad. This even included witch trials - apparently temperatures were so cold it made people hold trials for people being witches. I'm not even making this stuff up.
However, there were just a few problems with this. 1) The events are obviously cherry picked and focused on mid to high latitudes of the Northern hemisphere. 2) The cherry picked events were not compared to the temperatures where the events happened but to the temperatures in Central Greenland. 3) The temperatures in Central Greenland were from as single proxy, and other proxies from Greenland (and even the same ice core) show different temperatures. 4) The timeline was wrong - Andy May thought Years BP was before a.d. 2000, instead of before 1950 (he claims to be a paleoclimate expert!). And 5) the instrumental record was improperly plotted. When he posted this on Watts Up With That, these errors were pointed out to him, and he admitted to some of these mistakes - Kobashi's proxy differs from Alley's, he messed up Years BP, etc. But he still contends that in general his cherry picked historical events correspond with warm and cold periods in Greenland. This despite the fact that the event he calls "Starvation, extreme cold, over a third of the European population dies in 784" during the "European Dark Age" occurs during the warmest period in Kobashi's proxy reconstruction.[3]
When the evidence from Kobashi's paper became known to him, Mays had to switch out his cherry picked historical events to fit the new reconstruction. For instance, at a.d. 700 Alley's reconstruction shows a cool period which he associated with the deaths of a third of the European population. But Kobashi's reconstruction shows that time frame to be much warmer, here's his response. "The anomaly lasts 100 years and is precisely opposite of what the Alley reconstruction shows. I’m not sure how to interpret this. This was when Charlemagne unified what is now France and Germany into the Frankish Empire. It was also when the Franks (French) finally stopped the Moslem advance into Europe and held them in what is now Spain. Rashid, sultan of the Abbasid Caliphate, caused the caliphate to reach the height of its power at this time. Great empires are usually built during warm affluent times. I score one point to the Kobashi team." May's conviction about climate and history remains unchanged here. What changes is the historical events he cherry picks to make the "theory" work.
Now, how is this different from a Marxist looking through the newspaper to cherry pick confirming evidence for Marxism? Is this not precisely what we should be warning against? Andy May did not supply a single shred of evidence that climate change is not falsifiable. He did not supply evidence that climate science is a pseudoscience. However, we can see that May appears to lack skepticism with regard to his own claims. But there is ample evidence of May doing precisely what he criticizes climate science of doing - cherry picking confirming evidence in support of a pseudoscientific claim and failing to show proper skepticism in "proving ourselves and other wrong." As Feynman says, "The first principle is that you must not fool yourself and you are the easiest person to fool."
Note: I should acknowledge that, as important as Popper is for the philosophy of science, his work is not the last word on the scientific method. Popper's insights are valuable and will have a lasting and likely permanent impact on the scientific method. But falsifiability as I described it above can be a bit tricky. When theoretical prediction fails to conform to observations, in theory the failed test should cause us to revise the theory. But it's also valid to confirm that 1) the test was well-conceived, 2) the test was conducted properly, and 3) the observations/experiment isn't affected by another theory - in which case it could be that the theory responsible for the experimental values is wrong, not the theory being tested. Falsifiability is a product of the impact of logical positivism on science, which itself has issues. While Popper's insights certainly advanced the scientific method, we do not have to become "popperazi." In reality scientific inquiry and discovery is a bit more complex than Popper's theory.[4]
References:
[1] Andy May, "Climate Change and the Question of Science." Washington Examiner, Apr. 21, 2022.
https://www.washingtonexaminer.com/opinion/climate-change-and-the-question-of-science
[2] Andy May, "Climate and Civilization for the past 4,000 years." https://andymaypetrophysicist.com/2016/06/22/climate-and-civilization-for-the-past-4000-years/
[3] Andy May, "Comparing the Kobashi and Alley Central Greenland Temperature Reconstructions." Watts Up With That Blog.
https://wattsupwiththat.com/2016/06/25/comparing-the-kobashi-and-alley-central-greenland-temperature-reconstructions/
[4] "Beyond Falsifiability."
In this opinion piece, Andy May is claiming that Karl Popper's philosophy of science should cause us to see that climate change is unfalsifiable and therefore pseudoscience. Well, I think that's what he's trying to say but he kind of blunders it with a misunderstanding of Popper, climate change, and the scientific method. Let's start with this:
Both hypotheses and theories must be falsifiable. "Climate change" is not falsifiable. It is not scientific. Popper would describe "climate change" as pseudoscience since any weather event can be, and often is, interpreted as supporting the idea (much like the Marxist with his newspaper).
Now May is right that, to the extent we follow Popper, both hypotheses and theories must be falsifiable, but he has not made a case that "climate change is not falsifiable." And in fact, he hasn't even tried to make that case. Instead, he complains that people sometimes say that extreme weather events are all "presented as evidence of human-caused climate change--as proof that emissions of carbon dioxide are leading to an overheated planet." But who is doing this? Better stated, is this happening in such a way that would cause us to conclude that "climate change is not falsifiable" and therefore pseudoscience?
It should be obvious, I think, that you can't call climate science pseudoscience because of what politicians and the media say about climate science. If that were true, then quantum mechanics would be pseudoscience. Any evidence supporting this claim should come from the scientific literature, and in particular, peer-reviewed studies. Does the scientific literature present "tornados, hurricanes, mid-winter thaws, 100-degree days, cold snaps, droughts, and flooding creeks" as untestable (or unfalsifiable "evidence of human-caused climate change" or "proof that emissions of carbon dioxide are leading to an overheated planet?" Not in my reading of the literature, and May fails to provide a single example of this happening. Are we to just accept what he says here on his authority as a petrophysicist, paleoclimate expert and member of the CO2 Coalition? I hope not.
What does the scientific literature actually say about extreme weather events and climate change? A fair summary would be that as our emissions increase greenhouse gas concentrations in the atmosphere, we should see its effects on the climate system, including 1) an increase in global mean surface temperature and 2) the warming and acidifying of the oceans. These should in turn affect other aspects of the climate system, including 1) melting of glaciers and the shrinking of sea ice, 2) accelerating sea level rise from thermal expansion of the oceans and the influx of glacial ice, 3) increased uptake of CO2 by plants (greening), and 4) changes in weather patterns, especially extreme weather. Scientists also, through models and empirical methods, explore in what ways these extreme weather patterns will be affected by AGW. These include an increase in extreme heat, increased drought in areas prone to aridity (like in the American southwest) and increased flooding in areas prone to flooding. I could of course support each of these with evidence from the literature, but I think most people even moderately familiar with climate science would agree that the above is common knowledge, and I've interacted with the scientific literature here in other posts.
More importantly, every single one of these are testable theoretical predictions of climate science that demonstrate that climate science is indeed falsifiable. In the absence of stronger forcings pushing climate towards cooling, an increase greenhouse gases pushes climate towards warming. Is that observed? Yes. It should also warm the oceans. Is the ocean heat content increasing? Yes. It should also acidify the oceans. Is the average pH of the oceans decreasing? Yes. It should cause accelerating sea level rise. Is that happening? Yes. It should cause some "greening" due to the uptake of CO2 from plants. Has that happened? Yes. And it should cause changes in weather patterns, and in particular, extreme weather. Are we experiencing these changes? Absolutely. Extreme heat is far more common now than it was in the 1950s. Arid areas are experiencing more droughts and longer droughts, causing water shortages and increases in wildfires in these areas, while other areas are experiencing more flooding. More examples can be given here, but suffice it to say that, even if you disagree with me that these theoretical predictions have been confirmed with empirical data, it's absolutely clear that climate science contains theoretical predictions that can be tested against experience. Climate change is falsifiable.
What I think May may be getting at, if I give him the most charitable reading possible, is that it may appear that when unusual weather happens, that unusual weather is automatically blamed on climate change. If this is what he means, it's not "climate science" as a whole that is pseudoscience, but the aspect of climate science that attributes causation to weather events. He seems to think, whether you experience a heat wave or a cold snap, the event will be blamed on climate change. Whether you experience a drought or a flood, the event will be blamed on climate change. If every effect is automatically climate change, then this aspect of climate change can't be testable, and therefore it can't be scientific. But even this is not a fair assessment of the scientific literature (though some news stories do this). The scientific literature does not simply say we had bad weather, therefore AGW. Rather, the literature looks at how AGW theoretically should impact various regions, and that allows for testable theoretical predictions. So in the American southwest, an increase in temperature should deplete soil moisture by increasing VPD, which in turn should make droughts more frequent and severe. This is a testable prediction. And should it not be observed, theory can and should be revised to account for the reality. That's climate science functioning as a science. May provided literally no evidence of climate science itself operating as a pseudoscience.
What May does instead is wax eloquent about Karl Popper, though he doesn't get Popper quite right. He correctly points out that one characteristic that should be built into all scientific theories and hypothesis is testability or falsifiability. To investigate a theory of hypothesis, there needs to be a way to test it against experiment. So scientists develop theoretical predictions that can be tested against empirical evidence. Should it fail a properly conducted test, then the theory or hypothesis is wrong, and it must be discarded or modified in light of the new evidence. Any theory or hypothesis that can't be tested in this way is not properly scientific, which is why climate scientists go to such great lengths to test the theoretical predictions of climate science against empirical data.
But from here, May goes a bit wonky: "Popper draws a bright line between science and pseudoscience. The line is falsification. Pseudoscientific ideas cannot be falsified." The last sentence is absolutely wrong. There are many pseudoscientific ideas that can be falsified. It's clearly a pseudoscientific idea to say that the Earth is flat, but that is absolutely a testable claim - it is falsifiable, and it has in fact been falsified. Many pseudoscientific ideas are falsifiable. While it's correct to say that unfalsifiable claims don't qualify as scientific, many claims that qualify as being falsifiable are still pseudoscientific. Among falsifiable claims perhaps a good way of discerning pseudoscientific falsifiable claims would be 1) holding onto claims after they have been falsified or 2) rejecting claims without compelling evidence that have been confirmed by overwhelming evidence. As an example of the former, the cosmic ray hypothesis began as a legitimate scientific hypothesis, and it has been rigorously tested. It failed in part because trends in cosmic rays have been running counter to observed warming. There was absolutely nothing wrong with the scientists that proposed the hypothesis - we can learn a lot from proposing these hypotheses, even when they fail. What makes it currently pseudoscience is that those who hold to this view are holding onto a hypothesis that has already been falsified. Likewise, an example of the latter would be the rejection of the falsifiable claim increasing atmospheric CO2 pushes the climate system towards warming. This claim has been overwhelmingly tested against empirical evidence, and it has passed. It is not proven - there is a non-zero chance it could fail some test in the future - but the evidence for it is so strong it's pseudoscientific to reject it without also supplying compelling evidence overturning the evidence we have.
Can we find pseudoscience in larger discussion of climate science? That is, is there anything that qualifies as being comparable a Marxist searching for evidence confirming his theory in the newspaper and ignoring counter evidence? Absolutely. One of the more popular examples of pseudoscience is the claim that warm climates always correlate with thriving empires and cultures, while cold climates always correlate with societal collapse. This is a pseudoscience claim that fits exactly the example of a Marxist searching for confirming evidence for Marxism in a news paper. All you need to do is look for when temperatures are warm and find thriving empires and cultures that line up with those warm periods. Then you look for times when temperatures are cold and look for examples of societal collapse. Then you label the a graph showing these cherry picked events correlating with warm and cold periods, and there you go - pseudoscience at its finest.
It turns out, Mr. May has been doing this for quite some time. Back in 2016, he published a graph on his blog[2] showing "evidence" that warm periods are good for culture and cold periods are bad. This even included witch trials - apparently temperatures were so cold it made people hold trials for people being witches. I'm not even making this stuff up.
However, there were just a few problems with this. 1) The events are obviously cherry picked and focused on mid to high latitudes of the Northern hemisphere. 2) The cherry picked events were not compared to the temperatures where the events happened but to the temperatures in Central Greenland. 3) The temperatures in Central Greenland were from as single proxy, and other proxies from Greenland (and even the same ice core) show different temperatures. 4) The timeline was wrong - Andy May thought Years BP was before a.d. 2000, instead of before 1950 (he claims to be a paleoclimate expert!). And 5) the instrumental record was improperly plotted. When he posted this on Watts Up With That, these errors were pointed out to him, and he admitted to some of these mistakes - Kobashi's proxy differs from Alley's, he messed up Years BP, etc. But he still contends that in general his cherry picked historical events correspond with warm and cold periods in Greenland. This despite the fact that the event he calls "Starvation, extreme cold, over a third of the European population dies in 784" during the "European Dark Age" occurs during the warmest period in Kobashi's proxy reconstruction.[3]
Note: I should acknowledge that, as important as Popper is for the philosophy of science, his work is not the last word on the scientific method. Popper's insights are valuable and will have a lasting and likely permanent impact on the scientific method. But falsifiability as I described it above can be a bit tricky. When theoretical prediction fails to conform to observations, in theory the failed test should cause us to revise the theory. But it's also valid to confirm that 1) the test was well-conceived, 2) the test was conducted properly, and 3) the observations/experiment isn't affected by another theory - in which case it could be that the theory responsible for the experimental values is wrong, not the theory being tested. Falsifiability is a product of the impact of logical positivism on science, which itself has issues. While Popper's insights certainly advanced the scientific method, we do not have to become "popperazi." In reality scientific inquiry and discovery is a bit more complex than Popper's theory.[4]
References:
[1] Andy May, "Climate Change and the Question of Science." Washington Examiner, Apr. 21, 2022.
https://www.washingtonexaminer.com/opinion/climate-change-and-the-question-of-science
[2] Andy May, "Climate and Civilization for the past 4,000 years." https://andymaypetrophysicist.com/2016/06/22/climate-and-civilization-for-the-past-4000-years/
[3] Andy May, "Comparing the Kobashi and Alley Central Greenland Temperature Reconstructions." Watts Up With That Blog.
https://wattsupwiththat.com/2016/06/25/comparing-the-kobashi-and-alley-central-greenland-temperature-reconstructions/
Comments
Post a Comment