Communication with Steven Mosher regarding a potential collaboration

A while back, Steven Mosher made comments to the effect no skeptics are willing to work on a project to advance knowledge on certain issues.  I responded, suggesting that while I don’t know if I qualify as a skeptic, I would be willing to work on such a project.  This led to a discussion and an agreement to attempt a collaboration to examine the issue of the Urban Heat Island effect.

The collaboration failed.  I hold it failed due to poor behavior and bad faith on the part of Steven Mosher.  I repeatedly advised Mosher to take steps to address my belief, either individually or via mediation.  He failed to do either, and communciation ceased.  That was two months ago.

This collaboration was proposed and discussed in public.  This makes it a matter of public interest.  Given the discussion was limited, I conclude the public interest is also limited.  However, I believe that interest, combined with the perceived bad faith on Mosher’s part, justifies disclosing communication to allow the public to judge for itself why the collaboration failed.  I intend to do nothing to draw attention to it, but I will make it a matter of public record.

The beginning of the disagreements between Mosher and I began after he suggested I write a paragraph explaining the reasoning for the methodology I was proposing.  I wasn’t sure exactly what he was looking for, but I agreed, writing (my comments will be in blue, Mosher’s in black):

It turned out to be quite a bit more than a single paragraph.  It’s mostly stream of thought, and it might need to be edited into something more manageable before attaching it to an actual document.  Regardless, it should show my thought process clearly enough for you to figure out what you’d want included.  I’m not sure what you’re looking for, but I think this should suffice.

If you have any questions or concerns, let me know.

The “usual” approach to looking for a UHI effect has always been to examine the final results and see if one could find a UHI signal.  This is a weak test.  The failure to find a signal does not indicate the signal is non-existent.  The tests used could simply be inadequate.
Even worse, they could be inappropriate.  Previous tests have focused on whether or not the UHI effect biases linear trends in temperatures series.  Those tests have been predicated upon an assumption of what form a UHI signal would have.  They could not find a UHI signal that did not meet their expectations, even if it did exist.
A new method is suggested to address this problem.  Rather than look at final results for a signal, it is proposed we examine how the UHI effect could cascade through various data processing steps.  This is akin to how many systems are tested: One perturbs the data and sees what happens.
In this case, the hope is to create pertubations that mimic removing the UHI effect prior to data processing.  To whatever extent this is successful, one can directly test to see what differences are introduced by the existence of the UHI effect.  One can actually see how the UHI effect affects each step taken to create a final results set.
Moreover, this approach is highly modular.  If one questions any assumption or method used, changes can be made to test its effects.  By making every step distinct and its output accessible, any ideas or questions can be directly examined.  This can even be used to expand the scope of analysis by introducing any new signals one might be interested in (e.g. microsite biases).

Mosher’s response created several forks.  I’ll excerpt one as an example of what form the breakdown took:

The “usual” approach to looking for a UHI effect has always been to examine the final results and see if one could find a UHI signal.  This is a weak test.  The failure to find a signal does not indicate the signal is non-existent.  The tests used could simply be inadequate.”

The usual approach has been to test the following assertion

1. Urban sites  warm more than rural sites  and therefore the warming estimated by using all the stations will be biased high.
2. Nobody ever concludes that a failure to find the signal means it is non existent. quite the opposite everyone starts with the assumption that the bias is real.

This is a strange comment.  You say nobody concludes a failure to find a UHI signal means that signal is non-existent (even though some do) then claim the opposite is true.  However, your “opposite” refers to assumptions people start with, not conclusions they draw.  One could start with the assumption you list then reach the conclusion you say nobody reaches.  That means both can be true.  If both can be true, they cannot be opposites.

1. Even though some do?   This is easily settled.  Find the published science that says the signal is non existent.  What most people will say is that they cant find it, which is logically distinct from saying that it is non existent.
2. When you find somebody who says there is no UHI  or that the UHI signal is non existent, as opposed to NON DETECTABLE, then we can settle this. Its easy

Mosher, your entire response here rests upon misrepresentation after misrepresentation.  You’re simply making things up about what I say, and it renders any further discussion pointless until we can resolve this.  First, a thread began when I said:

The “usual” approach to looking for a UHI effect has always been to examine the final results and see if one could find a UHI signal.  This is a weak test.

This said nothing about what conclusions were drawn from the test.  It merely stated the test was weak.  Your response was a non-sequitur, saying:

2. Nobody ever concludes that a failure to find the signal means it is non existent. quite the opposite everyone starts with the assumption that the bias is real.

I ignored the fact you were using a red herring and engaged your response.  I pointed out your claim was nonsensical, and as a parenthetical aside, said some people do do what you claim nobody does.  Your response was to ignore the primary point I raised, that your argument was nonsensical, and instead focus on the parenthetical aside.  Even then, you failed to address what I said, instead saying:

1. Even though some do?   This is easily settled.  Find the published science that says the signal is non existent.  What most people will say is that they cant find it, which is logically distinct from saying that it is non existent.

I never claimed a published scientist concludes anything.  You altered my claim in a way that makes it more difficult to prove.  Not only did you fail to address my primary point, you created a straw man out of an unimportant aside.  That is a classic straw man argument. There is no way to justify it.

I do not believe a collaboration is possible when one partner constantly misrepresents what the other partner says.  I believe, at a minimum, both collaborators must be able to agree to what simple sentences mean.  Rather than attempt to reach such an agreement, Mosher proposed we drop all our disagreements.  Here is an excerpt which highlights the nature of that portion of the disagreement:

I made my objections. Im under no obligation to convince you. You did not accept my objections, so I accept your document as is.

Nothing in this exchange has hinged on the validity of any objections.  Your objections have never been an issue as they were total fabrications.  You can say you accept the document, but this entire exchange has been about the fact you’ve repeatedly misrepresented my proposed methodology.  Saying you accept the document means nothing if we cannot agree about what I say the methodology is.

The issue is you made things up about my proposed methodology in order to criticize it.  When I tried to resolve the disagreement of what my proposed methodology is, you constantly avoided meaningful discussion while creating further fabrications.

In the same e-mail, I explicitly stated my view of Mosher’s participation in this collaboration, saying:

I’ve specifically accused you of bad faith and misrepresenting my proposed methodology.  Saying you accept that document does nothing to resolve either concern.  You’re simply ignoring the issues I raise while focusing on a minor issue.

I do not believe you’re acting in good faith.  As I’ve said, that makes any collaboration impossible.  Given my accusations are based in specific examples, it is incumbent upon you to attempt to address my concerns, possibly by mediation.  If you choose not to, such as by suggesting it is my responsibility, no progress will be made.  In fact, until the time some effort is made at resolving the issues I’ve raised, this supposed collaboration is ended.

None of Mosher’s subsequent e-mails attempted to address the issues I raised in any meaningful way.  The last e-mail did offer some hope however, saying:

My understanding is that  you think it is my reponsiblity to pick a a mediator. That is fine.
I have a short list of three people. I’ll be contacting them.

Should you reneg I will understand.

He has not contacted me since.  I am unaware of who he had on his “short list” for potential mediators, and I have no idea whether or not he contacted anyone.  Whatever he may have done, I believe it is safe to assume there will be no further contact as there has been no follow-up in the last two months.

The full exchange is included in a separate document here (additional documents created are here and here).  Given it is in the public record, people can decide for themselves what they think of the participants.

I will note, however, Steven Mosher has often posted comments discussing how to interpret what people write.  This includes many comments discussing how to best interpret a comment when it has multiple possible meanings.  In addition to this, he has praised his ability to interpret comments, and he has openly admitted to “playing games” with people via the use of rhetoric, disingenuousness and strategic positioning.  That should be considered when examining accusations of blatant misrepresentations.

Advertisements

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