A look back at December 2022: “Duplicate” records

This is another very technical post, don’t bother with it unless you are obsessive over matching up numbers and pages in JFK ARC releases. This is my least favorite ARC subject by far, but if you want to understand the ARC review and release process in detail, it is something you can’t skip over.

This subject should also be required reading for people who believe there are major issues with ARC releases, like agencies hiding stuff or NARA not being honest. Will such people read any of this? I wonder.

The problem of “duplicate” records in ARC releases

I wrote a number of posts several years ago on the subject of “duplicate records” in ARC releases. In fact, this was the main subject of many of my posts on the 2017-2018 releases.1Interested readers can check here for most of these articles.

“Duplicate” may not be the right term for this problem in the ARC releases. In fact, the problems I discussed under this label probably include more than one issue. I will try to confine this note to one particular issue, however, which has come up AGAIN in the December 2022 releases.

The issue is this. In December 2022, NARA posted numerous sets of pdfs that all had the same RIF number, as shown on NARA’s excel file for the releases. This seems wrong: the RIF number for every document in the ARC is supposed to be unique.

As I have tried to explain before, when NARA posts a pdf for an ARC record on the ARC website, it does not mean that all the redacted text in the record has been released. It is not surprising to see pdf files for the same record posted in say 2017, 2018, and 2022. Each of these pdfs files releases further redactions, peeling off layers of redactions step by step. “Release” here thus refers to redactions, not to the record as a whole.

It is very odd, however, to see multiple pdfs posted for one ARC record on just one “release” date. Yet this has occurred again and again in the ARC releases. To make matters even more confusing, when one compares these “duplicate” records in one release, some of them seem to be completely different documents, and the excel file that NARA provides for the pdfs posted indicates that they may have a different number of pages and different original doc dates. Are RIF numbers really unique or not? Puzzling indeed.

The case of ARC 124-10167-10378

An earlier example of this can be found in the April 2018 releases. For this release, NARA’s excel file had four rows that all had the RIF number 124-10167-10378 (see here).

These rows were linked to four different pdf files on the ARC website. Following are the names and metadata from the excel file:

# filename pages est date
1 124-10167-10378.pdf 1 1976-01-22
2 124-10167-10378_1.pdf 5 1976-05-26
3 124-10167-10378_2.pdf 4 1976-02-27
4 docid-32989711.pdf 267 0000-00-00

The date of each document here is estimated based on a date shown in pdf file. Clearly the first three docs are different, yet each has the same RIF number, 124-10167-10378, in the NARA excel file.

The same oddity reoccurs in the December 2022 ARC releases. Searching again, we get the same result (the order is different, but I’ll use the same numbering as above):

# filename pages est date
1 124-10167-10378[c06716618].pdf 1 1976-01-22
2 124-10167-10378[c06716615].pdf 5 1976-05-26
3 124-10167-10378[c06716617] 4 1976-02-27
4 docid-32989711 267 0000-00-00

A this point, we will stop and try to figure out what we are looking at. The RIF data tells us these are pieces of FBI casefile number 62-116464. 62 often means an administrative file. An example of an important 62 file in the ARC is 62-109090, which is the file for FBI liaison with the Warren Commission. The file number for FBI liaison with the HSCA is 62-117290.

Browsing through the largest of these four files, it seems fairly clear that 62-116464 is an administrative or liaison file originally for the Pike Committee, and later for the House Permanent Select Committee on Intelligence. The “2ND NR 331 THRU 340” tells us that this is only a part of what must be a very lengthy casefile.

How lengthy? Well, it’s hard to say. There are 60 plus records in the ARC that come from 62-116464. All of them seem to be “Not assassination related” (NAR), which is the FBI equivalent of NBR, “not believed relevant” (to the JFK assassination). According to NARA’s James Mathis, the FBI did not fill out page numbers for NAR files.2See here, and there are indeed many blank page number fields for these FBI records. As we can now see, the big file we are looking at in 2022 is 267 pages. Some of the other 62-116464 file sections may be equally lengthy.

Going back now to the three small pdfs, what is their relation to the big one? It turns out that each of these is a piece of the big one. The one page file is page 120 of the main file. The five page file is pages 96-100. The four page file is pages 50-53. When we compare these pages to the main file, we can see that what is being redacted, and released, are the names of CIA officers. Thus the one page file has only one redaction, which turns out to be “Rogovin.” This is Mitchell Rogovin, who in 1975 was special counsel to the DCI. Why his name was redacted here, I do not know. In any case, it is now released. in fact, as far as I can tell, the main 267 page file has no redactions left in it.

The question still remains, however; why were the three bits of the main file cut off and separately released? I have a theory about that.

124-10167-10378 and the CIA “Transparency plan”

My theory is built on the CIA document list which is available here. This was released as part of the CIA’s response to President Biden’s October 2021 memo which ordered a further review and release of JFKARC records.3The memo is available here.

This document list is seldom cited in discussions of redactions remaining in the JFKARC. However, it contains much important, in fact essential, information on the redactions. As I have noted many times, the list covers all ARC documents that still redact CIA information. It covers both records released in full in 2022, and records that were still redacted after that date. If you want to know what was released and not released in 2022, it is therefore an authoritative source.

Looking at the list, one should perhaps not be surprised to find that it contains all of the three short excerpts from 124-10167-10378 which were separately released in 2018 and again in 2022, including the dates and number of pages. It also has a couple of new data columns, labeled “Case number” and “CADREREFID”. No idea what the latter means, but it turns out this is the number suffixed onto the RIF number; this combination was then used to produce the filenames for the 2022 pdfs posted at NARA.

On the other hand, the main file for 124-10167-10378 is not in the CIA document list at all. What this tells us is that the portions of the document with CIA officers’ names were listed as ARC records for CIA review, which makes sense. Only CIA can determine whether it is “okay” to release employee names. On the other hand, the portions of the main file which do not have such information (most of the file) were not subject to CIA review.

This implies that the CIA reviewed portions were treated as separate items, so the redactions (or releases) were also published separately.

Is this the reason for all the other cases where bits of one record were released separately from other bits of the record? Perhaps not, but I think this is very likely the case here. Releases, and redactions are officially recorded when published on NARA’s JFKARC website. This is one reason why NARA published even material that did not change from one release to another: to make a public record that it was unchanged.

Anyway, that is the theory that I have, and which is mine.