When Assumptions Replace Evidence In Climate Science

As discussed in the previous post, the Federal Judicial Center’s recently updated Reference Manual on Scientific Evidence contains a new chapter on Climate Science. [some emphasis, links added]

That chapter focuses on promoting the hocus-pocus of “attribution” studies that seek to blame every latest hurricane, flood, or drought on the human emissions of CO2, and thus on fossil fuel producers in particular.

In my post, I characterized the authors’ write-up of the methodology of these attribution studies as relying on “logical fallacy,” and as “double-talk and bafflegab.”

But I think that I inadequately articulated the nature of the fallacy. So I will try to correct that here.

The heart of the problem is that science is all about hypotheses being subject to empirical tests against real-world evidence. But the “attribution” studies and their methodology seek to evade that necessary step.

Instead, these studies claim to validate their attributions by reference to things like “physical understanding” and models that have not been empirically validated.

In other words, rather than using empirical evidence to validate a hypothesis, they use one hypothesis to validate another hypothesis. They have assumed the conclusion they want to reach. This process is sometimes called circular reasoning.

To be fair, when buried in the midst of enough confusing verbiage, circular reasoning can often be difficult to detect. That is not an excuse for the fancily-credentialed pooh-bahs who have signed off on this Manual.

If you aren’t up to the job of detecting circular reasoning, you are not qualified to contribute to a Manual like this one.

Here is the central portion of the key language in the new Manual chapter, as quoted from my previous post, with important words highlighted:

[A]ttribution involves sifting through a range of possi­ble causative ­ factors to determine the role of one or more ­ drivers with re­spect to the detected change. This is typically accomplished by using physical understanding, as well as climate models and/or statistical analy­sis, to compare how the variable responds when certain ­drivers are changed or eliminated entirely.

So, to make an “attribution,” we compare the event that just happened against our “physical understanding” and against our “climate models.”

Both of those things are our assumptions or hypotheses, not our proof. (The category of ‘statistical analysis’ cannot be evaluated without more information. ‘Statistical analysis’ of what? If it’s a statistical analysis of our current hypothesis against our previous assumptions, then it’s another example of circular reasoning.)

Thus, we have compared our new hypothesis to our preexisting hypotheses, and they match! Therefore, we claim that the new hypothesis must be correct!

The missing piece is some kind of empirical study that establishes a definitive relationship between rising global temperatures and the type of event that you are seeking to blame on CO2 emissions.

Remarkably, [empirical studies] do not exist for any form of extreme weather event, whether it be hurricanes, tornadoes, droughts, floods, or anything else.

A good compilation of studies failing to show any relationship between global temperatures and extreme weather events can be found in Steven Koonin’s book Unsettled.

Anyway, you know that there is no such demonstrated empirical connection, because if there were, the “attribution” advocates would cite those studies instead of trying to prove attribution with their own a priori assumptions.

Read more at Manhattan Contrarian