When Honesty Meets Stupidity

I was getting ready for bed, planning to go to sleep before one in the morning for a change because I have a doctor's appointment at eight. I decided I couldn't though, because I had one of those amazing coincidences in life that I just couldn't pass up.

You see, a few days ago I wrote a post about how a favorite of the skeptic movement, Richard Tol, secretly replaced one version of one of his papers with a new version to cover up an error he had made in the earlier version after it was publicly pointed out. That post was titled, "When Dishonesty Meets Stupidity" because I wavered on how much of Tol's actions could be attributed to dishonesty and how much of them could be attributed oo stupidity.

Well today, right before I was going to close my laptop, I happened to see a post at the blog Watts Up With That. The post accuses a bunch of people of being responsible for deception. For instance, it says:

Climate science is in a horrible state when you can pass off a bog-simple 12.6 year sine wave as a “harmonic solar component”. The journal, the peer reviewers, and the authors all share responsibility for this deception.

The thing is, there is absolutely no basis for that claim. The post doens't actually demonstrate the slightest shred of deception in the paper. As it turns out, the paper was completely up front and honest about what it did. So naturally, I had to laugh. The author of the post, Willis Eschenbach, had written a ton of verbiage and rhetoric to insult people all because he was an idiot. In other words, this was a case of "When Honesty Meets Stupidity."

Now, I was really wanting to go to bed, so I initially tried to just write a comment, not a blog post. It didn't work out, but the comment is somewhat enlightening as it explains:

What is going on with this site? Where in the world do you get off accusing people of deception for doing something they say they do? This post makes a huge deal of the fact the authors of this paper didn’t use solar data, but the authors never said they used solar data. They said they included a term in their model to try to account for the variability attributable to the sun. Whether or not you agree with their choice, they were perfectly up front and honest about what they did.

Heck, the paper explicitly discusses aspects of the nature of their choice of parameter for their model, pointing out solar forcing wouldn’t actually follow a sin wave, but stating they use a sin wave as an approximation to simplify the analysis. And this post calls them dishonest because… they didn’t use solar data? They told you they didn’t use solar data. You may not like that they didn’t use solar data, but they told you they didn’t.

You can’t accuse a person of dishonesty for doing an analysis in an up front, honest and perfectly clear manner simply because you don’t like it.

The authors wanted to create a model to try to explain the information present in sea level data. In doing so, they used some real data and assumed it had a particular type of relationship to the sea level data. They also included a modeled component. This component was supposed to represent the solar signal they hoped to find in the data.

Now, they could have used solar data instead of a modeled component to represent the solar component. They chose not to. Was that wrong? I don't think so. I think there are legitimate reasons one might do that, the primary one being it simplifies certain analyses.

However, that's not the issue Eschenbach raised. Eschenbach didn't just say, "The authors should have used real data instead of a modeled component." He said:

Climate science is in a horrible state when you can pass off a bog-simple 12.6 year sine wave as a “harmonic solar component”. The journal, the peer reviewers, and the authors all share responsibility for this deception. The study is not about “The solar and Southern Oscillation components in the satellite altimetry data” as the title claims. Iit’s not about solar anything. Instead, it is about fitting a sine wave to sea level data. That is false advertising, not science of any sort.

Not science of any sort? Eschenbach clearly has no idea what he's talking about. The modeled component the authors use is a reasonable fit for the actual data they used. Seeing how well it fits to a different type of observational data is certainly a reasonable thing to do. It may not be great science. It may not merit a published paper. But it is science.

And deception? What is Eschenbach smoking? The authors were perfectly clear about what they did. If you somehow didn't understand:

We include a solar term in the model because a large solar component was observed in various oceanic data sets, including the ocean heat content, sea surface temperature, and the sea level [Shaviv, 2008].

Meant it was a modeled component rather than observational data, though I'm not sure how you would think, you should certainly figure it out when a few sentences later the authors say:

The meaning of this phase will be discussed in the next section. Since the duration of the 23rd solar cycle has been 12.6 years, we force this period.

The authors couldn't force a period to exist in the component unless they were creating the component. And in the section titled, "Implications of the Solar Term," which you'd surely be interested in if you were interested in the "solar term" used by the authors, you'd see:

We have seen in the previous section that the sea level altimetry can be fitted with an empirical model that includes terms associated with the ENSO and the solar variations. The next step is to use the fitted parameters to derive constraints on the physical forcing terms. Here we shall do so for the solar forcing, which as we shall see has two primary components—a steric term associated with thermal expansion and a eustatic component, associated with water trapped in land reservoirs.

The above empirical fit assumed a harmonic solar forcing. Although it is only an approximation, it significantly simplifies the analysis. By describing the radiative forcing anomaly as a complex number: ΔFsolar(t) = ΔFsolar exp(−iωt), each component of the sea level can then be described with a complex amplitude. The phase will then describe a lag or lead relative to the solar forcing.

The section goes on to discuss in further detail the nature of the parameters of the modeled component called the "solar term." How much more clear could they have possibly been that this was not solar data? Whether or not you agree with their choices, the authors made it abundantly clear what those choices were.

Sure, you have to read more than the title and abstract of their paper to get the information, but... is that really what Eschenbach is complaining about? Is his pitching a fit and calling people dishonest because you have to read more than one paragraph to understand what's in a scientific publication?

No! Because if that were the case, he'd be a liar. Because he never once told his readers what I told you. He never once told his readers the authors acknowledged they used a modeled component. If he knew the authors did so, then he flat out lied to his readers. I don't believe he'd do that. I think he somehow looked at this paper, saw paragraph after paragraph of the authors describing what they did and somehow still failed to realize the authors were up front and honest about what they did.

Because he's an idiot. Or he's deluded. Or he's close-minded. Or something else. I don't know what. All I do know is this is ridiculous. Just two months ago, the proprietor of Watts Up With That, Anthony Watts, accused people of committing fraud because he didn't like the work they published. This is the e-mail he sent (to Peterson and his co-authors):

Dear Dr. Peterson,

This latest paper, Karl et al. 2015 is an embarrassment to science. It epitomizes president Eisenhower’s second warning in his farewell address about science and politics becoming hopelessly intertwined, and thus corrupted.

In my last telephone conversation with you, I stated (paraphrasing) that “I believe you folks aren’t doing anything fraudulent, but you are doing what you feel is correct science in what you believe is a correct way”.

After seeing the desperate tricks pulled in Karl 2015 to erase “the pause” via data manipulation, I no longer hold that opinion. You needed it to go away, so you prostituted yourselves, perhaps at the direction of higher ups.

This will be NCDC’s Waterloo, and will backfire on all of you terribly on the world stage. Take a lesson from Yamamoto’s own observation after he bombed Pearl Harbor. Take a lesson from what is on WUWT today.

How sad for you all.

Anthony Watts

When I said that was wrong of him, Watts turned around and denied used the word fraud. Seriously, Watts told me he hadn't used the word fraud:

Then changed his story to saying he hadn't ever accused them of fraud. He did so despite the e-mail being readily available to anyone interested, and despite me pointing out it clearly contradicted what he said. I won't go into what happened next as the point is Watts accused people of committing fraud and data manipulation without any basis, then when confronted, denied having done so and refused to try to make amends.

Now we have Eschenbach accusing people of deception without any basis on Watts's site. Will he deny having done so? I don't think so. One thing that makes Eschenbach different from Watts is he isn't a coward when it comes to things like this. I expect he'll keep insisting that a paper which repeatedly made it clear it was using a modeled component was dishonest for not using physical data.

8/14/2015 2:58 AM Edit: Willis Eschenbach has responded to me on his blog post. An interesting aspect of his remark is:

That is pure deception, Brandon. They are not using solar data precisely because the sea level is NOT in sync with the Sun as they falsely claim. If it were, they’d have used it. But they didn’t and that statement above is designed to deceive.

With this statement, Eschenbach claims to not only know the authors couldn't have reached similar conclusions to the ones they reached with real data (something his paltry work does not prove), but to know the authors are aware of that. Furthermore, he indicates that is the reason they chose to use a modeled component rather than use the real data. This is interesting as he not once said a word about the part of the paper which begins (and I quoted this above):

The above empirical fit assumed a harmonic solar forcing. Although it is only an approximation, it significantly simplifies the analysis. By describing the radiative forcing anomaly as a complex number: ΔFsolar(t) = ΔFsolar exp(−iωt), each component of the sea level can then be described with a complex amplitude. The phase will then describe a lag or lead relative to the solar forcing.

And then proceeds to describe how by using the modeled component they used, the authors can perform a series of analyses to try to draw certain conclusions. Given that discussion takes up something like 20% of the paper, that's incredible. To believe Eschenbach, we'd have to believe something like 20% of the paper was added in purely to cover up the fact the real data couldn't be used to reach the "right" conclusions.

But more interesting is the fact Eschenbach doesn't say anything about this. He doesn't inform anyone about these tests the authors performed. He doesn't tell anyone the authors have this potential reason for using a modeled component instead of real data. That creates a false impression where people reading what he writes wouldn't realize the authors had a stated reason for not using real solar data. Whether or not it was intentional, Eschenbach deceived his readers!

8/17/2015 4:48PM Edit: Willis Eschenbach has updated his article for something like the third time now, finally walking back from his libelous accusations that the authors were deceptive in doing an analysis they explicitly described doing in their paper. I don't know the full extent of the changes to the post as the Wayback Machine was down for scheduled maintenance when I went to check it to find a previous version of the post for comparison's sake, but the update notice says:

[UPDATE: Upon reading Dr. Shaviv’s reply to this post, I have withdrawn any mention of “deceptive” from this post. This term was over the top, as it ascribed motive to the authors. I have replaced the term with “misleading”. This is more accurate since it describes the effect of the analysis on the readers, and not the intentions of the authors. Dr. Shaviv and his co-authors have my apologies for my unwarranted accusation of bad faith.]

I don't know if my post here or comments there had anything to do with this change. Eschenbach may have made the changes because of them, or he may have made them because of Nir Shaviv's response. Or it may be due to some combination of the two. But this is from the last comment I made before he made the update:

Willis Eschenbach:

Brandon, thanks for the clarification. You still don’t seem to get it. Yes, they have attempted to justify the fact that they have not used real solar data. But do you actually think that if the actual solar data plus ENSO had been a strongly significant fit to the sea level, they would not have used the real data?

Okay, you are a coward, and a dishonest one to boot. This post portrays the authors as having deceived their readers by passing off a modeled component as real world data. Now, you acknowledge the fact they not only showed that their data was modeled, but tried to justify why they used modeled data instead of real world data. That means the entire basis for the claim of deception in your post was wrong, yet despite having backpedaled away from it, you won’t admit you were wrong.

You can disagree with the authors’ assumption all you want. You can say it is wrong. That’s fine. But there is nothing deceptive about making an assumption then carrying out an analysis based upon that assumption. The only person who has been deceptive here has been you, by pretending the authors were anything other than completely up front about what they did. And you’re still doing it here.

But please, stick with the libelous accusations based upon nothing but poor reading skills and amateur level analyses. I’m sure crying, “Liar!” every time you dislike a person’s approach will really help the skeptic cause gain traction.

Eschenbach didn't respond to me, but some time after this comment, he updated his post to retract his accusations of deception. I think that shows I was pretty on target with my criticisms. So was the comment I made before it (forgive the poor writing in it, I typed it on a phone across a number of red lights while driving):

Willis Eschenbach… so your accusations are based on nothing more than you being bad at reading? The authors devoted an entire section of their paper, which makes up something like 20% of the paper’s length, to analses tied to the fact their solar term is a sin wave. They explicitly say their choice of parameter is inaccurate to some extent but allows them to perform those analyses. It doesn’t require a “close reading” to understand that. It requires simply not ignoring one of the main components of the paper.

But even after pointed that section of the paper out to you, directly quoting it for you, you ignored it. Because… what? You just don’t read you criticize? You’re dishonest? Something else? I don’t know. What I do know is this is pathetic. There is no way anyone could actually read this paper with an open mind and fail to realize the authors used a modeled component for their solar term, and that doing so allowed them to perform a series of analyses which makes up a significant portion of the meat of the paper.

I don’t know how you failed to understand something so obvious, but it’s pathetic, and you owe the authors an apology. As well as the editors of the journal, whoever peer reviewed the paper and climate scientists as whole. Because quite frankly, this blog post is libelous, and if you were sued over it, I wouldn’t speak in your defense.

You, Anthony Watts and everyone else here can’t just go around accusing everyone of deception/fr*ud/lying willy-nilly just because you feel like it. It’s disgusting.

I pointed out Eschenbach's screw up within hours of his post going live. Eschenbach says I did so "quite unpleasantly," but whether I was nice or not has nothing to do with the fact it took him days to retract his baseless accusations. It took him days to admit he was wrong even though it was pointed out how obvious his screw up was within hours of his post going live. That's inexcusable.

3 comments

  1. Brandon, I am late to this thread as I was unaware it existed. I see that I was not as clear in my post as I imagined. Your defense of Shaviv is that in the paper he did exactly what he said he did. Let me see if I can explain why this is not sufficient by means of an example.

    Suppose I titled an article "Brandon Shollenberger is a member of the KKK". I start out the article by discussing the KKK and its history, and saying some general things about you. Then, halfway through the paper in a throwaway paragraph I say:

    "Since studying Brandon is difficult, we studied someone who looks a lot like Brandon"

    Then I went on to show how they can definitely prove that the individual that they studied is a member of the KKK.

    I'm quite sure that you would be upset with me, and you would say (correctly) that my article was misleading. Instead of studying you, I studied someone kinda like you, and I didn't include even one fact about you ... but I PUT YOUR NAME IN THE TITLE AND THROUGHOUT THE PAPER.

    But that is exactly what Shaviv has done. He has put "solar" and "the sun" throughout the paper and in the title ... but there is not one single scrap of data in the paper about the actual sun and its actual effects. Instead, they just fitted a sine wave of constant frequency and amplitude to some climate phenomenon, put in a tiny disclaimer, and called it a solar study. It is nothing of the sort ... but the paper has mislead lots of people. You can find claims around the web that Shaviv has proven that "the sun influences the climate" and the like, despite the fact that their paper had nothing to do with the sun.

    Now, I was totally wrong to describe this as "deceptive". Deception involves the intent to deceive and I have no idea what Nir's intent was. That was wrong, and I apologized for it.

    However, just as in the example I cited above about the notorious KKK member Brandon S., it is most assuredly misleading.

    So I stand by what I said. Describing the fitting of a sine wave of constant amplitude and frequency to climate data as being about the sun is as deceptive as studying someone other than you and claiming that it actually has something to do with you.

    I hope this clarifies things, as in general I'm a big fan of your thoughts and ideas.

    Best regards,

    w.

  2. I don't think you follow me on Twitter, so to repeat what I said there, it looks like I'm supposed to be offline for about a week due to some medical issues. Bedrest is boring though, so I've still checked a few things. I saw this comment in an e-mail alert, and I felt I needed to respond promptly to say something:

    Willis Eschenbach, please read this paper. Actually read it. It was clear you hadn't bothered to when you wrote your post about it, and I'm still not convinced you have. If you had, I cannot understand why you'd write:

    Suppose I titled an article "Brandon Shollenberger is a member of the KKK". I start out the article by discussing the KKK and its history, and saying some general things about you. Then, halfway through the paper in a throwaway paragraph I say:

    "Since studying Brandon is difficult, we studied someone who looks a lot like Brandon"

    Because that is nothing like what was done in this paper. This paper discussed the fact it didn't use solar data numerous times, discussed the reason for having done so and argued the choice wouldn't impact their results.

    Whether they were right about the choice affecting their results or not, they were perfectly clear about what they did. Anyone who read more than just the title and abstract of the paper would have been aware of what they did. The only people who could possibly be "deceived" by this paper like you claim the authors were hoping are people who didn't bother to read the paper.

    There is nothing to clarify about your position. Your position is perfectly clear. It's just wrong. You didn't bother to read the paper before writing your post. Perhaps you skimmed the paper. Perhaps you only read certain sections. I don't know. What I do know is you specifically claimed the authors of a paper attempted to deceive people by passing off a modeled parameter as real world data, despite the authors spending something like 20% of their paper discussing this very issue. When I pointed this out, you defended your claims, highlighting parts of the text you felt supported your accusations without addressing the significant portions of the paper that contradicted your portrayal.

    So please, read the paper. If you do, you'll find it spends a great deal of time discussing the thing you claim the authors tried to deceive people by hiding. That will show you you are wrong to continue with this ridiculous portrayal the deception was only disclosed in a single line "in a throwaway paragraph."

    Or don't. If you prefer not to read the papers you criticize, that's your choice. You'll just be mocked for it and perhaps even called dishonest for your continued repetition of obviously false claims. Either way, I'm exhausted now, and I'm going back to bed. If you write another comment like the ones you've written to respond on this subject (here or at WUWT), I'll probably not respond. At this point, the falsity of what you say is obvious enough nobody needs me to highlight it any further.

Leave a Reply

Your email address will not be published. Required fields are marked *