Archival Research

written textual records (documents)

Documents (written textual records) at the National Archives and other venues

Searches for records usually fall into of of two categories:

- the client requests records regarding a general subject, or
- requests specific records that have been previously identified.

    Whatever the case, the responsive records are either original hardcopies, microfilm copies or the records have already been scanned and posted to the NARA website (in which case, the client can simply download them). If the records are hardcopies or microfilm, my job is to locate them and make digital copies. This sounds easy enough, but there are often unanticipated pitfalls. 

Sequence for locating records

    Usually the client comes to me with a clear subject and will have already queried various online databases, such as those listed on the National Archives homepage or its newer Online Public Access page.  In order to double-check, I usually repeat the client's queries, before ordering the records. Sometimes the client has already communicated with a NARA archivist.  If so, it can be worthwhile for me to talk with the archivist and go over the project in detail. Archivists, especially those at the National Archives are often overworked - having to deal with dozens of people asking about a variety of unconnected historical projects, and also having to respond to emails and telephone calls.  Accordingly, NARA archivists are not always accessible.  However their insight can really speed up a search process. 

    In the case of the records of some agencies, such as those of the State Department (depending on which year they were created), many records can be found on the NARA website - and more of these are being released in digital format all of the time.  In such cases I refer the client to the exact location of the files in the NARA website. The one problem with the digital images of records on the NARA website is that they mostly  black and white reproductions and the conversion processes sometimes misses items written in certain colors- as well as the color of the paper - which can make a real difference for State Department documents.

    If the records are hardcopy written records or in the form of microfilms, in order to request them, one must determine their exact location in the "stacks." Unlike a library, these "stacks" are stored out of view. The details of the records are often not listed online but can be found in a "Finding Aid," a loose-leaf folder listing the contents of boxes by Record Group, Entry, Box and often Folder (rg#e#b#f#). This is turn leads one to a "MLR" (Master Location Register) - usually another loose-leaf folder that details the exact location of the records in the stacks by Record Group, Stack Area, Row, Compartment and Shelf.  The location of the records may be narrowed down to a single folder in a box or to dozens of boxes containing several square feet of records, depending on the subject.

    Once in a while clients come to me with citations they have found in the literature with apparently clear location information - the exact record group, entry and box number with the "exact" location in the stacks.   But there are pitfalls: as more records are being merged into the stacks at the archives, whole record groups or sections thereof can be re-boxed, thus invalidating the old locations. In many cases there are Finding Aids available that reveal the concordance with and location of the cited records in the new boxes.  In some rare cases, records have been re-boxed twice, thus greatly complicating the search. Worse, sometimes there are no Finding Aids or MLRs for that specific subject and one must order series of "likely" responsive boxes.

    Records are ordered and obtained by filling out a "Reference Service Slip"  stamped and signed by an archivist.  These are then processed several times a day (weekdays only) and one can pick up the boxes - usually placed on a cart - within a half an hour or up to to two hours, depending on how many researchers are waiting for them.

    A maximum of 24 boxes can be checked out on a single cart, and a total of two carts per researcher are allowed.  Only two record groups can be checked out at a time - but cannot be ordered at the same time.  Once the researcher has signed out the boxes, they can be taken to a table in the research room where they can be opened and their contents examined.

    A single box can take from a few minutes to go through - or more than day, depending on the complexity of the subject, the readability of the records, and, believe it or not, the thickness of the documents. Many of the records from the last century and before are printed on onion skin paper (click here for an extreme example) which is less than half the thickness of regular paper, which means one needs twice as much time to go through them. Other records are printed on both sides.

    While I look through these records, I often use a smart phone camera to photograph or video the results and then email images to the client in real time. These photographs or videos are not designed to be used as a final product: rather they can be used to vector in on the responsive records. For an example of a photograph, click here and for a video, here.

    Once records are located, digitized copies must be made and sent to the client.

Digital copying

    Responsive records are either photographed with a high-resolution Nikon D90 camera (about 10 MB) or scanned with the Plustek A360,1 the fastest professional large format flatbed scanner commercially available (I am one of only two NARA professional researchers who owns one). The resulting digital files are then adjusted and edited in Adobe Photoshop Lightroom and then converted into easy to print text-searchable PDF2 files in ABBYY FineReader 10 and saved as Adobe Acrobat 9.5 Professional files. These files are then transferred to the client over a secure Internet connection, such as YouSendIt - often the same day.

    Depending on the amount of staples to be removed, I can usually digitize some 600 - 800 pages as unedited and unadjusted JPGs per day. However, if the client wants color-adjusted text-recognized PDFs fewer images can be processed.


1 None of the scanners costing below $800.00 come close to the speed of this scanner, which can cycle a complete scan of an  8 x 10" page (landscape) in about 1 second, a full fifteen times faster than  the other scanners on the market.  It will also handle documents as large as 12 x 17".

2 Searchable text-recognized PDF files are created after the image has been converted into a PDF file using an OCR (optical character recognition) program. Such files are searchable by many search engines, such as Windows Desktop Search and Google Desktop Search or the search tool that in included with Adobe Acrobat Reader. These search tools do not work with handwriting and can be imprecise, especially with old documents written with manual typewriters.