Sunday, November 22, 2009

The word "trick"

People seem to often be focusing first on this word "trick" used in a Phil Jones email:

In one e-mail from 1999, the center's director, Phil Jones, alludes to one of Mann's articles in the journal Nature and writes, "I've just completed Mike's Nature trick of adding in the real temps to each series for the last 20 years (i.e., from 1981 onwards) and from 1961 for Keith's to hide the decline."

Mann said the "trick" Jones referred to was placing a chart of proxy temperature records, which ended in 1980, next to a line showing the temperature record collected by instruments from that time onward. "It's hardly anything you would call a trick," Mann said, adding that both charts were differentiated and clearly marked.

Mann is certainly right: "trick" is a word scientists use all the time to mean a clever way to refocus a problem or transform it somehow to make progress. If I had a dollar for every time I heard a professor say "Now here's a trick you can use," I'd be able to buy a Lexus.

An elementary example of what might be considered a trick is converting from Euclidean (x,y) coordinates to polar (r,θ) coordinates where a calculation simplifies. A more complicated example is dimensional regularization in quantum field theory, where, since some observable factors are calculated to be infinity, you instead do the calculation in 4+d dimensions and then in the end let d go to zero, and the answers are finite. (No, there is no good mathematical basis for this, as a mathematician will tell you, but as a physicist will tell you, it works.)

Skeptics who are focusing on this particular word just do not understand some of the inside language scientists use.

22 comments:

rhhardin said...

I don't think anybody's worried about "trick."

It's the way the work is being defended from investigation.

Any trick would be discovered very fast without the email dump, if the work could actually be checked over.

That seems to have been their worry.

So the pretty certain feeling is that the work won't in fact stand up to investigation, and they know it.

Model builders are not like other scientists in their orientation.

It's like they have the education but not the curiosity that would hold their interest in real work.

rhhardin said...

My impression from the passage as I recall it was that "trick" meant what you said (technique, not, ie, trickery), but the trick was done to keep the end from turning down or something like that.

That is, the result was fudged by a certain technique.

That the result is sensitive to the technique is the problem.

Now the technique has to be defended, whatever it was.

David Appell said...

> That is, the result was fudged
> by a certain technique.
>
> That the result is sensitive to the > technique is the problem.

No, that's not right. The technique is simply described in the Wash Post article I linked to.

Define "fudged."

David Appell said...

> Any trick would be discovered
> very fast without the email dump,
> if the work could actually be
> checked over.

The work was "checked over" -- via peer review, and by subsequent researchers.

As Mann explains in the Post, the "trick" has been in the scientific literature for years.

rhhardin said...

"Fudged" means picking the technique that gives the result you think is true.

Anonymous said...

David, I think the point in that email is the phrase "hide the decline" rather than the word "trick", which can have entirely innocent meanings.

I've been thinking all weekend, and I do wonder how the schism in the body politic can be fixed. We need to get back to a situation where different points of view are respected. The most troubling part of these emails to me is the declaring of certain people as "enemies" to be neutralised.

Do you have any comment on the emails about eliminating journal editors or the "joke" about beating people up?

Slowjoe

Anonymous said...

The problem, David, as the e-mails make abundantly clear, is that the "peer review" check was nothing more than a filter to keep out anti-global warming folks. There is a string of blunt instructions on perverting the process at all of its stages. Appalling.

Dano said...

the e-mails make abundantly clear, is that the "peer review" check was nothing more than a filter to keep out anti-global warming folks.

Aside from the fact that Stevie Mac was little more than a purposely time-wasting tick,

Can you explain the filter shown in the e-mails surrounding the issues of the Douglass et al. papers? You know: the fraud by Douglass, Christy, . Pearsona and S. Fred Singer? Did they keep these folk out?

Thank you so much in advance.

Best,

D

David Appell said...

> The problem, David, as the
> e-mails make abundantly clear,
> is that the "peer review" check
> was nothing more than a filter
> to keep out anti-global warming
> folks.

I don't see that at all. First of all, no scientist/peer-reviewer has the power to "keep out" anyone. The journal editor makes the final decision about the paper. And there is usually more than one peer-reviewer, sometimes three or more.

Besides that, places like CO2science.com love to list all the peer-reviewed papers that support their position. How did M&M get published if gatekeepers are keeping people out? Lindzen? I've seen other lists of peer-reviewed papers that supposedly disprove pieces of climate science. How did all these get published?

David Appell said...

> Do you have any comment on the
> emails about eliminating journal
> editors or the "joke"
> about beating people up?

Yes -- scientists are human, and some of them are at times not even be nice people. A few are real crumbs.

Do you think this kind of competition doesn't go on in high-energy physics, in nanotech research, once in AIDS research, etc? And nanotech has practical applications worth billions. Scientists usually guard their secrets closely, so they can make more advances and get the credit and shut out their competitors. Science is hardly some big happy playground where everyone shares and is nice to one another. It's a competitive field like all fields and industries out there.

If I thought a journal editor was making editorial mistakes, suppressing good science, or publishing bad science, I'd too wonder what was going on and if he might be replaced.

If anything, these CRU emails show that scientists are human and have all the traits of humans, both good and bad. Did you really expect differently? If someone looked through all of your emails, what kind of person would they conclude they are? Could they pick out excerpts that pain you as a jerk? They certainly could with my emails, and I suspect with anyone's.

Besides, if you take the inverse of all of these emails, I doubt they would look any different from a collection of emails sent around to skeptics.

nod said...

I second or third the point that the issue is with "hide the decline", not trick. Trick is so trivial to explain it should go without saying.

I don't think there is sufficient context in the email alone to determine if "Hide the decline" indicates a dishonest action. Although I will add it seems unlikely to me on the face of it that a scientist would email their colleagues to tell them they had just done something dishonest. If that was the level of play I would have expected a lot more emails than just this one.

I think "hide the decline" can be justified, but it's not clear from the email that it is. Then again neither are the accusations of fraud which skeptics are making of it.

The decline of course was false -temperature didn't decline over the past few decades.

If you are trying to plot temperature over time and you have confidence in the paleo reconstruction up to the divergence, then it wouldn't be fraud to hide the false decline with the actual increase.

What would have been wrong would be to hide the decline simply to hide the divergence because it was felt it harmed confidence in the paleo reconstruction up to the divergence.

So which one is it? The email doesn't say.

It did seem improper to me to just stick the instrumental record and paleo reconstruction together. They should have been overlaid.

Well this was back in 1999 or 2000? if this is the best the skeptic can find in 10 years of email then I am not convinced. If manmade global warming is all a UN conspiracy then I am deeply unimpressed at the PR abilities of the UN vs the skeptics!

Anonymous said...

"I have been attending climate conferences for years now -- a couple each year, as many as I can on my tiny freelance earnings, and hardly all of them -- and have attended talks by scientists locally. I've questioned them in person and on the phone and via email. And I have never once seen any canonical scientist "behave badly." On the contrary, everyone I've talked to or communicated with -- including many of the big names -- have all behaved professionally and politely and responsibly, without fail. Without fail." - David Appell

I find this old comment of yours "oddly cheering".

Anonymous said...

You are trying to gild the lily. Parse the evidence properly! "Hide" means hide, means hide, means hide, means hide. It does not mean anything else.

It demeans you to provide excuses for what is manifestly scientific fraud.

nod said...

shortened Anonymous: Hide means hide means hide means hide means...scientific fraud.

Anonymous said...

I agree with the person who said that it was NOT the word "trick" (which we all use - perhaps to point out how legitimately clever we perceive ourselves!), but the subsequent "hide" which is the indicator in the email. Scientists endeavor to "un-hide". Explain away the "hide".

But better still, forget the English language and its silly, poorly defined words such as trick and hide. We have computer code - which is 100% unambiguous. Either the data is manipulated, or it is not. Why do you suppose the whistle-blower(s) included the code?

David - it's over.

Dano said...

David - it's over.

Yup.

The smoke screen is clearing and there's no fire. Except for some pants.

Over indeed.

Best,

D

Anonymous said...

There seems to be a confusion about what is (or is not) being hidden. The “Hide the decline” issue is not hiding the cooling (or leveling off if you prefer) of the last 10 years - it is about the “divergence problem”.

If, to make a more familiar example, at your home you update from a mercury thermometer to a digital one, you would likely leave both in place for a period of time for comparison – likely concluding sufficient equivalence.

In the climate study, the “old thermometer” is very odd – tree rings. Clearly, not just temperature, but precipitation, nutrients, etc., all influence tree rings. Extracting temperature is extremely difficult. This makes a careful study of the overlap period ESSENTIAL. Specifically, the post-1960 instrument record needs to be compared to the post-1960 tree rings. If they are in good agreement, we trust the "old thermometer" more. Otherwise, otherwise. The instruments show an increase, the tree rings: a decline (a divergence) that is then "hidden".

GW alarmist should rejoice at the instrument increase. But they then hide the divergence. If the tree rings do not correctly measure post-1960, we can’t trust them pre-1960 – they might miss the Medieval Warm Period, etc. Apparently, getting rid of the MWP is the essential step. Very BAD SCIENCE.

turboblocke said...

Coming a bit late to this party: but as 1998 was a super hot year and the e-mail was written in 1999, what decline do the deniers expect it to be referring to?

clarkbeast said...

Dr. Jeff Masters over at Wunderground has a clear explanation of what "hide the decline" means:

"The paleoclimate data after 1960 show a bogus decline in Earth's temperatures that does not agree with what modern thermometers have been measuring, due to a well-known variation in tree ring thickness as a function of time, referred to as 'the decline.'"

The age of a tree ring would seem to be a known variable that needs a "trick" to be corrected for (or "hidden") so that you're comparing apples to apples in the data. Nothing nefarious whatsoever.

http://www.wunderground.com/blog/JeffMasters/comment.html?entrynum=1392

Anonymous said...

Why are you repeating this straw man? Why does every defender of the faith repeat this straw man?

It seems to be some kind of viral meme.

Most people interested in this debate know exactly what "trick" means in this context, and the word "trick" is not the problem. The problem is the precise nature of the trick, and the purpose to which it is put. These are spelled out in the email.

The nature of the trick is to splice together data from tree rings and data from the temperature record, as if they are the same thing.

The purpose of the trick is to hide the fact that, from around 1960, tree rings in the sample used diverge from the temperature record -- tree ring thickness declining while temperature record rises -- for reasons speculated about but unknown.

The reason to hide this is to hide, or at any rate downplay, the fact that tree ring data is demonstrated by the fact of divergence not to be a reliable proxy for historic temperatures.

If the tree rings can diverge from temperatures for unknown reasons, any graph of tree rings purporting to be a graph of temperatures is unreliable, and indeed meaningless. If this fact is not hidden, it will be obvious at first glance to any alert person seeing the graph.

Thus, the trick is a deliberate attempt to hide the fact that the graph is bad science. Thus, it is a clear case of scientific fraud.

Another thing implicit in the email is that the field is happy to publish work that very obviously falls short of the basic standards of good science, if the findings of the paper can be made to appear to support certain a priori conclusions.

This same group, we can see from the emails, will go to great lengths to prevent anything being published that reaches conclusions contrary to AGW theory, and will even boycott a journal that has the cheek to publish a paper that criticizes their work.

These guys are not scientists. They are bigots with ideas so fixed, they are unable to see how far they have gone wrong.

Joe Goodacre said...

The issue is not with the word trick.

The issue is that the trick/technique was done to 'hide the decline'.

Scientists are meant to be objectively searching for the truth, not activists.

Attempting to hide data which shows a decline suggests activism. For an activist to dress up as a scientist and intentionally lead people to a conclusion that they would not hold, had all other data been presented (declining and all) is fraudulent as the word is generally understood.

It's curious why more people such as yourself who supposedly value science fail to recognise how behaviour such as this is so damaging science. People don't have the time to read every article themselves and every paper. Science is based upon trust. To the extent that people see to be manipulating results (i.e. hiding the decline) then that trust goes.

David Appell said...

Joe: Do you know what the word "decline" refers to? A decline in what?