I’ve used a typewriter since at least 1967 and computerized paint programs since the mid 1980’s. I guess in birther terms, I have 72 years experience as a forensic document examiner. But actually, I am not a lawyer, a doctor, a typewriter expert nor a computer graphics guru, but my position as a blogger entitles me to write on these topics anyway; and while I am not a forensic document examiner, I was privileged to listen to one on the Reality Check radio program and I once attended a talk given by a (reformed) professional forger. I served for two years on the Fraud Prevention Committee of the National Association for Public Health Statistics and Information Systems, the national vital records association. As a result of all of my experience, I know enough not to be fooled as easily as some.
One of the things I learned is that forensic document analysis is a scientific process and when properly done, uses the scientific method. The way a real forensic document examiner normally determines if a questioned document is likely to be a fake is to compare it with similar documents that are unquestioned—the more “similar” the better, and the more examples available1, the more certain the document examiner can be about the results. Two things muddy the waters when looking at Barack Obama’s long-form birth certificate, the lack of real forensic document examiners, and the lack of real documents—both questioned and unquestioned.
Ideally, one would want to look at the original birth certificate in the files of the State of Hawaii and to compare it to other unquestioned contemporary original documents, preferably typed on the same typewriter in the same time frame. That’s obviously not going to happen for a variety of reasons and it increases the uncertainty of a scientific analysis greatly.
A more realistic scenario would be a comparison of the certified copy of Barack Obama’s long-form birth certificate to a certified copy of another certificate from the same time frame and preferably typed on the same typewriter. This approach is also impossible to perform presently because of lack of access to such documents, and the general refusal by the Hawaii Department of Health to issue certified copies of other long form birth certificates. Lack of access to these documents increases the uncertainty of the analysis.
A fallback scenario would be the comparison of the images of documents released by the White House to images of other contemporary certified copies of long form birth certificates from Hawaii, preferably from the same time period and in the best of all possible worlds, typed on the same typewriter. The current lack of good images of contemporary certificates hinders this process. Here are some examples that are unquestioned (click to enlarge):
A brief comparison of the documents above and the White House birth certificate shows that the President’s certificate shares many common characteristics with the unquestioned documents, so much so that it is either authentic, or informed by authentic certificates. That latter is the birther approach, claiming that the White House birth certificate was “digitally constructed” by taking bits and pieces of other certificates and assembling them.
One can ask the question, “is the White House PDF a scan of a paper document or not?” I think that the scientific method could be brought to bear on that question and an answer gotten as to what is most likely. That would involve comparing the White House PDF file to a similar unquestioned file of a scanned paper document. Here’s where birther analysis is unscientific and why one questions the validity of birther results (and in particular that of the Arizona Cold Case Posse). The key word is “similar.” For comparison purposes the “similar” file should be an image of a similar document produced in a way similar to how the White House PDF is represented to have been produced. The Cold Case Posse introduced a “control” document, but it was not an authentic document, but something they printed themselves that superficially looks like the White House file, but has been optimized and smoothed multiple times. The second problem is that the process and equipment the White House used is unknown. One piece of software used by the White House was part of the Mac system: Quartz PDF Context. To my knowledge no birther analysis, and certainly not the CCP) has used this software.
One of the birther objections to the White House PDF is a “halo” effect seen around the letters. Not having a comparable example, this objection cannot be evaluated. However, I did notice that in Adobe software (not used in the White House process so far as known) there is a suggestion that haloing is an issue, based on this dialog from the “Optimize Scanned Document” function in Adobe Acrobat 9.0.
So what is “Halo Removal” for if scanned documents don’t have this problem? Perhaps the kind of haloing the birthers find in the White House PDF is different from other kinds. There’s no way to come to that conclusion because birthers are comparing apples and orange. As a result any conclusions based dissimilar documents and processing, such as we see with the CCP, is invalid and unscientific.
Birthers raise other arguments that don’t rely on the assumption that no paper long form birth certificate exists. I’ll talk about one of those in “Forensic Document Stuff (Part 2–Typewriters)”.
1Addional samples help differentiate between significant and random variation.