Sunday, October 31, 2010

(S)He who refuses to do arithmetic is doomed to talk nonsense

So I realise that I'm sort of contradicting myself in going on about this, but I think it's important to be clear about it, especially as it's more-or-less on my home turf and I certainly don't want Michael Tobis to think he's out on a limb (or perhaps more likely, for his readers and especially critics to think he's out on a limb). So let's make this plain:

The "Italian Flag" analysis, at least as implemented by Judith Curry, is incoherent nonsense. She displays no clarity of thought on what the categories actually mean, or whether there is any workable calculus underpinning the whole thing. Going back to the original documents that she cites, it looks like it might have be supposed to have something in common with Dempster-Schafer theory, but I'm not sure about that and it's certainly not compatible with her usage. While it might be possible to reverse-engineer some semblance of sense into some of her statements regarding it, they are mutually incoherent.

And to those who are trying to change the subject and claim that these numerical details don't matter, you are dead wrong. If her accusations of IPCC errors regarding uncertainty are to have any credibility whatsoever, it can only be through her own analysis demonstrating such errors. So far, as Stoat also points out, she has run away from basically every challenge, merely saying things like the following, on a lengthy collide-a-scape comment thread:
"The fact that the climate blogging community doesn’t get what I’m talking about makes me pretty worried about the intellectual foundations underpinning the whole argument."
Well yes, Judith, when you find that everyone else is out of step, it is probably appropriate to worry about the intellectual foundations underpinning your whole argument. But somehow I don't think you meant that.


Anonymous said...

I'm curious what you think of her interpretation of the IPCC's uncertainties, specifically, not to do with the flag, but to do with the model's aerosols, solar, early 20th century forcings, etc. IMO, non-expert as it is, I think she incorrect. And I think it is important because, like yourself, she would like to make changes to the IPCC and science review in general. I had this uniforming exchange with her on her blog. What do you think?

Steve L said...

I think that, if you're going to claim that somebody else is quantifying uncertainties incorrectly, then you'd better come up with an argument that is logically correct and numerically workable. The fact that she seems incapable of the latter (even a biologist like me can tell) should definitely make you wonder about her competence in concluding the former.

Anna Haynes said...

There are people who say we should just ignore her. But Nature's not ignoring her, Scientific American's not ignoring her - and at least one journalist (probably plenty more, given their pack nature) is eager to give her yet more airtime.

So it's necessary to call this stuff out, when it promises to continue being a source of public confusion.

David B. Benson said...

Gone emerita, as the saying is.

James Annan said...

Hi gryposaurus,

It seems to me that you are entirely right in your exchanges there.

Curry's argument, such as it is, is based on yet another fundamental misunderstanding. The studies she refers to (AR4 are following an entirely separate strand of research from the production of the 20th century simulations using GCMs. Some of the work *uses* the *outputs* of the GCM simulations, but the GCM simulations absolutely were not tuned using these inverse methods. I find myself in increasing agreement with Eli, and it looks like a clear case of Dunning-Kruger. Her discussion is totally jejune (until she does her usual trick of running away).

Rocco said...

Anna Haynes: She should be ignored. Right now there is huge demand for scientists saying anything as long as it is "very not the IPCC". So she has no real incentive to be coherent.

When scientists respond to her, it is unlikely that anything useful will come out of it, and they will be portrayed as "the inquisition" for criticising her (huge demand for that narrative, too).

We thought her getting her own blog would help, but no. I guess the only venue left is the scientific literature.

Anonymous said...

Thanks James.

I find it unfortunate that Judith continues to pursue this method of dealing with IPCC uncertainties with modelling. Whatever those uncertainties are, it does not seem at all like she is giving a useful account of it. Even more, I imagine the new calls from the blogesphere will be filled with words like 'kludge' and 'tuning' and 'circular logic' instead of having an mature examination at what the models can tell us.

Unknown said...

Anna Hayes, if you're still reading this thread, come over and play in the devil's garden, where this journalist won't ignore here.

But lucky you, I'm not ignoring you either:

Marion Delgado said...

Sourcewatch spelled "Collide-a-scape" accurately.

    1. ↑ Keith Kloor (2010-04-23). An Inconvenient Provocateur. Collide-a-scape.