Home | The most likely author of the Heartland Institute climate strategy memo?

The most likely author of the Heartland Institute climate strategy memo?

A computer analysis may suggest the document's most likely author

By Shawn Lawrence Otto | Feb 23, 2012 | Comments (0)

A lot of fuss is being made about the provenance of the Heartland Institute's climate strategy memo (PDF).  Heartland says it's a fake, although as I showed several days ago, it is, if anything, a milder version of the information contained in much greater detail the apparently authentic Fundraising Plan (PDF). 

So why claim it's a fake?  One explanation is because the document was not intended for the entire board, but for a select few, as it itself says.  This could prove internally embarrassing for Heartland's senior staff and could possibly damage relationships with board members of the second class. 

Another is that there is something as-yet undiscovered about the document that is incriminating or embarrassing in some other way.

But the most obvious reason would be, as Heartland's communications director Jim Lakely told me, because it really is a fake.  Gleick says it was anonymously mailed to him.  Perhaps this was by a whistleblower, or perhaps it was by an disgruntled insider.  Or perhaps it was a honeypot - a sweet trap designed to compromise or discredit Gleick by getting him to write about it, while Heartland could trumpet how it is not authentic - in which case it would seem Gleick turned the tables by posing as a board member and requesting - and receiving - a cache of authentic Heartland documents.

At this point this is all conjecture.  There is simply not enough data.

Anthony Watts, one of the climate deniers the Heartland documents show is slated to receive tens of thousands of dollars in support from Heartland this year, has suggested that the document was forged by Gleick after he got the other documents.  This seems improbable, considering that the other documents are far more inflammatory. 

But Watts makes another suggestion which, being a writer I found intriguing: perform stylometry and textometry to see if Gleick really did write the climate strategy doc.  Watts even helpfully suggests an open source java app called JGAAP that purports to do this.

I decided to try it out. 

 

Methodology

The program works by entering a document by an unknown author and then it compares it, using various user-selected analyses, to a document(s) by known authors.  Here's what I used:

Unknown Author Document:
1. The disputed Heartland Institute climate strategy memo. (docx)

Known Author Documents:
1. Peter Gleick's mea culpa from HuffPo (docx)
2. Peter Gleick's previous HuffPo piece on climate and water. (docx)
3. Joe Bast's responses to Gleick' mea culpa (Bast is the president of the Heartland Institute) (docx)
4. Joe Bast's piece criticizing Sara Reardon's piece in Science. (docx)
5. Heartland Staff's Section 6 of the Fundraising Plan. (docx)
6. The disputed Heartland Institute climate strategy memo - as a control.

I supplied links to the above documents so you can verify my methodology and attempt to replicate my results.  As you can see, I copied the above six documents into word documents for uniformity but did not alter the language in any way.

The program analyzes according to several possible methods, which you can choose.  It lists results by low score to high, low being the most probable author of the unknown document.

The program and my methodology may be subject to flaws. I may have typographical errors in my documents that could influence the results.  I may not have chosen the best methods of analysis.  The documents I selected may not be a large enough representative sample of the respective writings of the various authors.  I may not have chosen a broad enough selection of authors.  The program may contain logical or mathematical errors.  I would encourage others to attempt to replicate, critique, and perform other analyses.

 

Results

Of the six author choices of: The Memo Itself, Peter Gleick 1 & 2,  Joe Bast 1 & 2, and Heartland Staff, based on the above criteria, here are the scores JGAAP assigned for most likely authorship of the climate strategy memo under just a few of the many analyses the app can perform:

Heartland Strategy Memo.docx
Canonicizers: none

Analyzed by Nearest Neighbor Driver with metric Camberra Distance using Character 2Grams as events
1. Strategy Memo 0.0
2. Joe Bast 3.2756019350109358
3. Heartland Staff 5.861152017670673
4. Peter Gleick 7.631295386657848
5. Joe Bast 10.572152376359865
6. Peter Gleick 11.883756639524362


Analyzed by Nearest Neighbor Driver with metric Camberra Distance using Word 2Grams as events
1. Strategy Memo 0.0
2. Joe Bast 2.599109316906122
3. Heartland Staff 6.170704701235744
4. Joe Bast 9.570177815725275
5. Peter Gleick 13.307560177813828
6. Peter Gleick 13.695029284565496


Analyzed by Nearest Neighbor Driver with metric Camberra Distance using Word stems as events
1. Strategy Memo 0.0
2. Joe Bast 3.8820363096787065
3. Heartland Staff 7.695783407036921
4. Joe Bast 12.653793919968829
5. Peter Gleick 14.734167804512905
6. Peter Gleick 16.420190717794636


One possible flaw not considered in the above methodology is that the program could be attributing too close of an authorship match to Joe Bast and Heartland Staff because the strategy memo contains a sentence that also appears in the authentic Fundraising Plan: "Principals and teachers are heavily biased toward the alarmist perspective."

I next crontrolled for this possibility by deleting the sentence in question from the climate strategy document, which I resaved as Climate Strategy Memo 2 (docx) and reran the same three analyses, using the original, full strategy document as one of the possible known authors, and got the following results:

Heartland Strategy Memo 2.docx

Canonicizers: none
Analyzed by Nearest Neighbor Driver with metric Camberra Distance using Character 2Grams as events
1. Strategy Memo 1.938705726184192
2. Joe Bast 2.9968217577188563
3. Heartland Staff 5.5472178990613275
4. Peter Gleick 8.156321472803725
5. Joe Bast 9.863690024204885
6. Peter Gleick 11.241849893833598


Analyzed by Nearest Neighbor Driver with metric Camberra Distance using Word 2Grams as events
1. Strategy Memo 4.183330096592706
2. Joe Bast 5.2640798702672384
3. Heartland Staff 7.577255315445771
4. Joe Bast 8.593442340043726
5. Peter Gleick 12.32193237311855
6. Peter Gleick 16.31481171381239


Analyzed by Nearest Neighbor Driver with metric Camberra Distance using Word stems as events
1. Strategy Memo 3.188360519110196
2. Joe Bast 4.783646657279247
3. Heartland Staff 9.110105530159261
4. Joe Bast 12.56463219477823
5. Peter Gleick 14.959700479974499
6. Peter Gleick 16.735394841607917

 

Conclusion

According to the above six analyses, which may contain unknown errors, the most likely author of the climate strategy memo is Heartland Institute president Joe Bast.



Get Shawn Lawrence Otto's new book: Fool Me Twice: Fighting the Assault on Science in America, Starred Kirkus Review; Starred Publishers Weekly review. Visit him at http://www.shawnotto.com. Like him on Facebook. Join ScienceDebate.org to get the presidential candidates to debate science.

 


Tags: Antiscience, Climate Change, Politics, Science, Media, Technology, Crime, Cool Research

Blog Rules

Comments (0)







Allowed tags: <b><i><br>Add a new comment: