It’s been a couple of weeks since I reported in with an update on my work at the APS, and what an eventful two weeks it has been. Buckle your seatbelts, folks.

Having transcribed roughly 10% of the 372 pages of the post office book, my supervisor Scott Ziegler and I realized that the transcription process was moving pretty slowly (see my last update on some of the practical issues I’ve run into while transcribing the book). Although I’m not engaging in a transcription process as detailed as would be required for bibliography or textual analysis,1 I still have to decipher lots of puzzles and figure out how to best render them in a form that can be analyzed as pure data. This is very time consuming, requiring up to an half-hour per page. At this rate, I wouldn’t have been able to finish the transcription anyway. Instead, we decided that the bulk of my remaining time on this project should be used primarily for research, interpretation, and thinking creatively about the application of data visualization techniques.

Pennsylvania Gazette (May 9, 1754)
A 1754 issue of Benjamin Franklin’s newspaper, The Pennsylvania Gazette. Image in the public domain.

I have made a lot of headway on these fronts, and I’ve really enjoyed the research process. As I’ve mentioned before, I’m most interested in the human beings behind the data, and figuring out what I can about them and their lives from the postal records. I began my approach to figuring this out by identifying the most frequent Post Office users from my transcribed spreadsheet of data so far. My hypothesis was that “the regulars” would be merchants, so I dug into contemporary newspapers- local titles like the American Weekly Mercury, The Pennsylvania Gazette, and the Woechentliche Philadelphische Staatsbote as well as issues of the New-York Gazette and the Boston Post-Boy, all of which are handily digitized and available through Readex’s America’s Historical Newspapers database.

It soon became evident that based on the information so far transcribed, my hypothesis was correct. Of the 21 most frequently occurring names of individuals, 15 were explicitly described as merchants, and several more were or had been ship captains involved in the import of goods. Two were lawyers. Having established that this tack might be fruitless in terms of uncovering anything not already known, I took a step back and reevaluated my source material, asking: what exactly can this data tell me? I came up with a short list:

  1. Most obviously! Who received mail, and where it came from, geographically
  2. Who picked up their mail, and who didn’t (The postal system at this time required payment upon receipt of a letter or package, and this is recorded in the ledger.)
  3. Who was receiving official government mail (This is the only mail that did not require postage, and is indicated as “free” or “paid” in the ledger.)
  4. The weight, and therefore number of sheets of paper, of the average postal transaction (For instance, the majority of items weighed less than 10 pennyweight, which by my math translates to just over half an ounce)
  5. The relative amount of postal interchange between Philadelphia and other cities (For instance, the amount of mail sent to and received from Boston was much more prolific than that to/from Williamsburg.)

Feeling slightly more optimistic about the utility of the data, and therefore more inspired to do creative things with it, I set about plugging the geographic coordinates of locations named in the ledger into the Google Maps API to get a sense of where the most mail seemed to be coming from. This are screenshots of my result, as a cluster map:

This slideshow requires JavaScript.

So far, so good. But how else could I depict the information in the ledger book in a meaningful way? In assembling short biographical sketches of the colonial Philadelphia Post Office power users, I found references to the relative locations of their shops and homes- I say relative, because I hadn’t really ever realized that our current system of house numbers and street addresses is a product of the USPS. This ledger predates individual addresses and so locations are described as “in Second-Street opposite the Friends Meeting House.”2

Yikes, how frustrating, huh? But wait a minute… I remembered that I’d come across a 1762 map of Philadelphia annotated with the names of individual shipping wharves and important places. It would be very cool if I could demonstrate the varying locations of the shops and homes and triangulate that with the location of the Post-Office, at the time inside of Benjamin Franklin’s printing office. I wasn’t sure how to go about it, but I am lucky to be surrounded with a lot of very smart people at APS. I got to talking with one of those folks, Steve Marti, the Digital Humanities Fellow, who has been working on a series of interactive visualizations of Eastern State Penitentiary admissions records. Per his suggestions, I plan to use LeafletJS to layer the historic map on a modern one, and pinpoint the locations of shops and homes of frequent postal service users to see if physical proximity to the Post-Office may have played a role in who utilized its services.

This map-building effort will occupy the remaining few days of my fellowship, in addition to some data-tidying and writing a brief summary of my findings.

I don’t know if I’ve yet addressed the issue of silences that I ruminated on in my last post and have spent countless hours considering over the course of the fellowship. Last week, thinking about this very issue, I remembered Laurel Thatcher Ulrich’s oft-misquoted statement that “Well-behaved women seldom make history.” Ulrich’s call to arms was not an appeal for anarcha-feminism as commonly interpreted. Instead, it was a plea for historians to start writing about those everyday, common women whose traces are all over primary source materials but whom are often ignored in pursuit of other stories. If I can’t fill those academic silences with scholarship in a field that is not my specialty, I can at least make the primary sources that feature these overlooked historical actors more accessible, more obvious, less easy to disregard. Maybe someday someone will uncover Jubbah’s story, because this volume is now freely available. I’d feel pretty okay with that.

Total hours: 112


1. Just in case you’re curious about “true” transcription and/or bibliography, I recommend a review of the short, practical volume An Introduction to Bibliographic & Textual Studies by William Proctor Williams and Craig S. Abbott. I personally find the process very satisfying and engaging, but it’s not appropriate for the purposes of this data set in this context!
2. Address of Joshua Maddox, Esq., per the Pennsylvania Gazette, March 5, 1754.

I’ve finally returned to work at APS following two weeks of travel and one of illness (boo!). Luckily I made quite a bit of progress this week and last, and I feel like this project is now moving along at a steadier pace. In fact, beginning the week of July 17 I will be increasing my weekly hours at APS to facilitate quicker completion of the project.

I was delighted to finish digitization of the volume, and the high-resolution scans are now available to researchers in the American Philosophical Society Digital Library. This is the first time that the book’s contents have been available to researchers outside of the physical confines of the APS, since this particular volume is not reproduced in the Benjamin Franklin PapersThe next phase of the project is transcription into a structured spreadsheet, which I got started with last week.

The transcription process has instigated consideration of how to structure the data I’m transcribing to make it the most useful and accurate representation of the messier, sometimes complex manuscript version. Besides making decisions about standardization (“Wm.” or “William”?) and guessing at abbreviations (“Do.” = “ditto”?), I’ve also made sure to record page numbers and URLs of the digital pages so that they can be referred back to later. This is a time-consuming process and I anticipate that I will be working on this phase of the project for a good chunk of time. I’ve also been drafting some blog posts for APS as well as tabulating the time that it takes me to perform certain tasks for the purposes of contributing to the Digital Library Federation’s Digitization Cost Calculator Day of Data.

One of the more interesting things that I’ve been thinking about as I work through the data is what this book could actually reveal besides things like the cost of postage in 1748 or the volume of mail sent to Annapolis for any given week. As always, I’m interested in the human element, and I want to know who was using the postal service. Then, as now, there would have been multiple methods for sending letters and packages- some cheaper, some quicker, some more reliable.

My initial hypothesis was that it would have been mostly businesses or merchants who used the system the most. I can’t be sure yet, but I’m beginning to reconsider that idea. I’ve found a few women named by their given names (i.e. not “Mrs. James Wilson” or the like). I’ve also found some rather cryptic references to addressees with only one name- like “Jubbah” and “Sabinah,” shown below. Could these people have been servants or enslaved and receiving letters (which they would have had to pay for upon receipt)? I’ve been thinking a lot about how silences in historical conversations are sometimes made based on flawed evidence collection. A lot of what information is gathered depends on what the historian is looking for and where they decide to look for it- and if their eyes skim past Jubbah because Mayor Edward Shippen’s name appears below it, that’s a deliberate silence.

Jubbah: The man, the myth, the legend?

I will attend the Keystone DH conference Wednesday through Friday, and probably blog about it, and then I will up my hours beginning July 17th.

Hours for last week: 14
Total hours: 41

I’m settling in at APS and enjoying the work and workplace atmosphere. My second week was much like the first, involving lots of digitization and image treatment of the aforementioned Franklin Post Office Book, but this time with the addition of Technical Difficulties- an unavoidable but still odious aspect of doing digital work.


I’d hoped to have the digitization phase of the project finished this week, but my progress was slowed by issues with hard drive storage space, diminishing image resolution quality, and fickle image treatment software. I found myself scanning portions of the book over and over again while troubleshooting and desperately pleading with the computer to function for at least an hour before requiring a reboot. I’m pretty sure it looked something like this:

In a panel at the James A. Barnes Graduate Student Conference this spring entitled “The Importance of Being Digital: How Can Graduate Students Help Build Digital Bridges?,” I advocated for digital humanities work in the undergraduate classroom as a way to acclimating students to ‘failing’ as part of the learning process. It’s extraordinarily rare to succeed on the first try of anything, but digital projects tend to highlight this facet in a sometimes obnoxiously prominent way. This week’s experience was a good reminder for me to be patient, take a deep breath, and try again when things just don’t work the way they “should.” I’m learning! It’s a process! What else can I do!

The other thing I’ve been working toward this week is the bibliographic essay that will be due for the in-class portion of this summer practicum. I’m really rather at a loss as to what topic I want to focus the assignment on. If this will eventually serve as a lit review or be incorporated into my thesis, I think I’ll need to focus on some aspect of digital humanities methodology. My supervisor at APS, Scott Ziegler, has been a lot of help in finding articles and talking this through, and I keep thinking back to the hack vs. yack debate after our discussions. I’m not completely sold on using it as a topic, because I’m not 100% on board with the ideas implicit in the debate to begin with. Do I really want it to figure into my thesis?


In other news, I was inordinately excited when a friend mentioned that she saw I’m listed on the APS website as one of the Digital Humanities Fellows. Huzzah!

I’m heading to Chicago for the Victorian Society in American Summer School, so I won’t be logging any new fellowship updates for a couple of weeks until I get back. I will likely have some reflections on the Summer School here.

I leave you with a lovely view of the Thomas Jefferson Garden at the APS Library, with the APS Museum in the background. I feel so lucky to spend the summer here.
This time from the other side of the fence!

Hours for this week: 14
Total hours: 27

The start of June marked my first work week as the first annual Martin L. Levitt Fellow at the American Philosophical Society Library. The fellowship is a joint venture between the American Philosophical Society Library and the Temple University History Department to honor Dr. Martin Levitt, an alumnus and faculty member of Temple’s graduate history program and longtime APS librarian. I’m honored and excited for the opportunity to work at this esteemed institution for the summer of 2017.

My project will focus primarily around creating open data from one of Ben Franklin’s sensibly named “Post Office Account Books.” Here’s the finding aid entry:
Screen Shot 2017-06-05 at 9.50.52 PM

The first step has been to digitize the book, a surprisingly lovely and tidy volume written in manuscript. I will likely finish this task in the next week. It entails scanning the book in high resolution using an overhead book scanner, then performing light image editing on the resulting scans to make them as legible as possible. The leaves and binding of the ledger are in quite good condition, and Franklin used a non-acidic ink, so the scanning has gone smoothly without the need for conservation work or convoluted repositioning measures to hold the book open. I have done work like this before at the Maryland State Archives, but I hadn’t worked with materials so closely linked to American colonial history before now. It has led to some mildly interesting discoveries, like the doodles shown below.

There are best practices for scanning and editing procedures (image resolution, file formatting, file naming conventions, etc.), but as for now, I haven’t encountered much clash between public history theory and practice, so I don’t know if I can write about that as of yet. This phase of a digitization process almost always happens “in private”- and especially so because the volume in question is a monetarily and historically valuable one. I expect that as the project moves forward I will encounter some situations in which theory and practice interact more directly- perhaps in the transcription and data organizational phase, which is intended to suit the needs of the public that will eventually consume and use this data.

Hours for this week: 13
Total hours: 13