Homework #2: Finding Ads in POTH

If you prefer, you can download these instructions in PDF form.

Runaway slave advertisements from nineteenth-century Texas appeared in newspapers that have been digitized. That is, they have, like all digital representations of analog sources, been partially digitized. The Portal to Texas History at the University of North Texas contains full-page images of many nineteenth-century newspapers, together with metadata about the newspapers themselves and OCR text for each newspaper page that makes it possible to search for text.

But these newspapers have not been digitized so as to provide metadata or descriptions at the level of individual articles. That means, to paraphrase Daniel J. Cohen and Roy Rosenzweig, part of the information visible to the eye (i.e., information about when a new article or ad begins and ends) has been lost (or at least not digitized) in the process of the newspaper’s "becoming digital."

This presents a problem for researchers, like us, who are interested in a particular kind of article—runaway slave advertisements. In this homework assignment, you will engage in the practice of digitization by looking through page images from one year of the Telegraph and Texas Register, identifying advertisements pertaining to runaway slaves, and inputting some basic metadata about the ad into the collaborative spreadsheet that you used in Homework #1. In the process you will also learn to pay attention to the "interface" of a search database and gather new information about how acts of resistance or flight by enslaved people were represented in primary sources.

Objectives

  1. To gain familiarity with how one major digitization project has decided to produce and share digital objects.
  2. To generate new questions about the kinds of information contained in runaway slave advertisements and how they changed over time.
  3. To help complete a complete database of runaway slave advertistements found in one of Texas’s longest-running nineteenth-century newspapers.

Before You Begin

This homework assignment will require you to spend time using the Portal to Texas History, so begin by watching this introductory video about the project:

Also spend some time browsing the site and looking through the help guide for the site, particularly the one on using newspapers. Think about how the site is organized and what sorts of searching and browsing are possible (or not possible) with the user interface provided. Run a few searches about something that interests you, and click through on the results. Get a feel for how the site "works," spending at least 10 minutes on this before proceeding.

Now head over to this lesson from the Programming Historian website about downloading records. You’re not actually going to be "programming" for this assignment or doing any downloading; you should only read the first three major sections of this lesson: "Applying Our Historical Knowledge," "The Advanced Search on OBO," and "Understanding URL Queries." These sections give you a tour through the Old Bailey Online, whose search query interface is broadly similar to the Portal to Texas History. Pay particular attention to what the lesson shows you about "query strings." Then go back to the Portal and run some more searches, noting how the values in the URL query strings change as you navigate through the site or run different searches.

Now you are ready to proceed to the homework assignment.

Finding Ads

Step 1: Find Newspaper Issues

Each of you will receive an email from me assigning you a year (or the equivalent number of issues) from the Telegraph and Texas Register that you will be responsible for reading in search of runaway slave advertisements.

Your first task is to figure out how to perform a search (or modify a search URL) so as to pull up (in "date ascending" order) all of the issues from the newspaper in your time period.

Here’s an example of what such a "Search Results" page looks like for the 1843 volume of the Register:

Search Results showing all available issues from 1843

Once you have a page that you believe shows the first page of all the results from a search for issues of the newspaper in your assigned timeframe, tweet that URL directly to me @wcaleb with the course hashtag so that I can check the URL and make sure you have found all the relevant issues. I have to approve this URL by a reply tweet before you can continue.

Step 2: Find Runaway Ads

Now you will be ready to go into each issue and look for ads. You’ll click through to the "Read this Newspaper" tab of each issue, and then click on "Zoom (Full Page)" so that you can magnify the image. Use the arrow pages at the top to flip through the various pages (or sequences) of the issue. Even though the ads are most likely to appear on pages 3 and 4, make sure that you at least run your eyes over every inch of every issue.

This will take time so start early. I recommend that you time how long it takes you to get through one or two issues following the steps below, so that you can plan your schedule accordingly.

The ads will come in different formats, and may have very different amounts of information. Some of the ads will be posted by subscribers who are seeking to find a slave who has runaway. Other ads will be posted by sheriffs or others who have captured a slave and are seeking the legal owner. If it looks like a runaway slave ad to you, or just looks like it has to do with runaway ads (e.g., a notice from the newspaper about how to submit a runaway ad, or an item about the different graphics used in ads), you should go ahead and enter it into the spreadsheet. Right now we just want to identify items of interest, so better to cast a wide net than a narrow one!

Once you’ve found an ad, you’ll need to enter it on the Google spreadsheet of runaway ads already collected. Be sure to carefully follow these instructions when you enter:

  • Before you start entering, find the appropriate "sheet" by looking at the tabs at the bottom of the window. Each year has its own "sheet" or tab, so find the one that belongs to your year.
  • As shown in the labels at the top of the sheet, you should list the year, month (as a number) and day (as a number) of the issue on each row.
  • Also include the full citation (which you can copy and paste from the top of zoomed page at the Portal to Texas History):
Full Cite Information for the Issue, as seen on Zoomed Page

Full Cite Information for the Issue, as seen on Zoomed Page

  • To generate a "permalink" URL that can be copied into the permalink column, first use the "Zoom" feature to enlarge the ad and center it in your browser’s viewing window. Make it as large as you can while still keeping all of the ad within view. Then click on the "Permalink" button while zoomed in on the ad.
Permalink button

Permalink button

  • If you think you recognize the ad as one you have seen before in a previous issue, identify it as a reprint by placing an asterisk in the final labeled column. You can use the other blank columns to the right to make any helpful notes to yourself (for example, by noting the names as a way of helping you to remember which might be reprints).

Finally, if you go through an entire issue of the paper and find no runaway ads, make a single row that indicates the date of the issue, and then type "None" in the "Full Cite" column.

Step 3: Reflect on Findings

After you have finished looking through all of your assigned issues, return to the JSON gist that you submitted for your first assignment and notice what pieces of information you found significant about the three ads you looked at then. In the process of going through your year of newspapers, did you notice new kinds of information that you had not seen before? Are there name/value pairs you would add to your JSON if you were doing it again? Was there anything about the newspapers (either in the ads or in the surrounding material) that surprised or interested you?

For the final step in this assignment, leave a comment on this blog post answering at least one of these boldfaced questions. You may use a non-identifying pseudonym as you make your comment, so long as you let me know which pseudonym you used.

Summary

To recap, successful completion of this homework requires:

  1. A tweet to me, with the course hashtag and the URL to the first page of search results from Portal to Texas History containing all the available issues of the Telegraph and Texas Register in your assigned time period.
  2. A completed tab in the Google spreadsheet that documents all of the ads contained in the paper in the time you were assigned.
  3. A brief comment reflecting on what you saw in the newspaper.

Points will be deducted from the assignment if the above technical requirements are not met or if the work contains numerous typographical errors, as well as for blog comments that do not seriously engage with the questions asked and reflect a thoughtful encounter with the newspapers you saw.

As in the first homework assignment, you can always take to Twitter if you need help, but in keeping with the academic integrity policies for the course, do not get someone else to do the work for you and be sure to acknowledge any pointers or technical assistance you received—in this case by noting it in your blog post comment.

8 Responses to Homework #2: Finding Ads in POTH

  1. One of the things I noticed while going through the runaway advertisements for HW2 is that they tend to follow very similar formats. Most of the ads contain similar sets of information such as the slave’s name, physical features, and descriptions of clothing. For these ads that follow a similar format, it would be easy to put the information into a standard set of JSON name/value pairs. Occasionally though, I came across an ad that contained some really unique descriptor. For example, one ad for a slave named Joe included the detail that he “was in the Alamo with his master when it was taken”. This sort of unique information is interesting and should definitely be recorded somehow, but doesn’t fit a set of JSON name/value pairs as easily.
    Something else I noticed from my year is that a subscriber put out separate ads for the same runaway at two different times during the year. The first ad for Gumby first appeared in June, and then later in December another ad for Gumby appeared (not a reprint but a new ad). Although this is just one example, studying ads from one newspaper across time could potentially reveal trends about slaves who ran away and were caught multiple times.

  2. Chronicling runaway slave ads throughout a year allowed me to notice trends that were not available to me through my previous exposures. The ads in the newspaper were primarily centered around runaways who had already been caught and were kept in jail until their masters reclaimed them. Many of the months had sparse to no advertisements, but the beginning of 1841 was extremely full of relevant advertisements. These ads were heavily reprinted, which indicates a certain lack of success in finding the owner of a particular slave. Questions that arose from this were: how long did ads to reclaim runaways run? Was there a procedural length, and if so, what happened to the slave following the given time period? From the mere reprinting of an advertisement, I became invested in a specific slave Charles who was waiting to be reclaimed by his master. The last sentence of his advertisement (repeated so many times!) stated the alternative to reclamation as “the law will be enforced.” What did this enforcement specifically entail? So, although I feel like many of the categories of information qualified relatively well with the JSON strings, the chronological unity of researching an entire year allowed a more global perspective on the slave advertisements and how they relate to the academic community. Asking more questions is always a good sign of history!

    • Re: your question about the law, it turns out there was a law passed by the Republic of Texas in February 1841 that required the capture of runaways to be advertised by county sheriffs for a set period of time, after which the runaway would be sold at public auction.

      Give that this law required the publication of ads like the ones you are finding, one question this raises is how legislation might have affected the patterns of publication and republication in the newspapers. That is, private calculations by owners were not the only things driving the incidence of ads in the paper.

  3. I would add to my JSON code the following fields:

    – Nickname. In one of the runaway ads I found, the slave’s “real name” and “claimed name” (what he goes by) were both listed. Cursory research tells me that some slaves changed their names after being freed, and presumably some may have masqueraded under a new name after running away to hinder detection. Thus, other ads might list multiple names for a runaway or suspected runaway and this is important information to capture.

    – Liberty. This could take on the values, for example, “Slave,” or “Suspected slave.” In one of the ads I found, there was uncertainty whether a woman living with a runaway was a slave herself or free.

    – Physical description. Most runaway ads had some sort of physical description of the slave, and while the information is usually not in a standardized format that makes it easy to analyze to answer broad questions (like eye color: brown/blue/black, weight: heavy/thin/medium), it is still important to capture.

    – Last know location and original (runaway) location. Storing this information makes possible spatial representations. Geo-coordinates would be necessary to map locations, but storing text-locations is the first step in that process.

    – Type of ad, for example “for claim,” or “missing.” Some of the ads were posted by sheriffs who want to return the slave to its original owner. Following this for “for claim” ads, we would want a field for the finder’s name, as well.

    One ad for purchasing slaves surprised me. Its implication is self-evident:

    The undersigned wishes to purchase two or three likely young negroes, for whom the highest price in cash will be given. I also want a good strong wagon…

  4. The patterns of runaway slave ads I found in this newspaper in the year 1845 was very different than I was expecting, and I am interested in seeing how major events during that year in Texas possibly could have affected runaway slave ads (such as the annexation of Texas). During the beginning of the year, I saw at least one runaway slave ad per newspaper issue, and most of them were reprints of the same few ads. Midway through April, the ads suddenly became more sporadic, with many issues lacking even one ad. Many of the newspaper issues in June through mid-September did not contain ads, and the ones that did all had a reprint of the same ad. There were no more actual runaway slave ads for the rest of the year, but there was a news article about finding a runaway slave in one of the November issues of the newspaper. Then, at the end of December, someone paid to have an ad looking for “four good strong Negro men” to hire for work at a steam mill reprinted in a few issues. It was interesting looking through a whole year’s worth of newspapers, because I expected there to be ads more consistently through the year. I am curious to see how the patterns of 1845 compare to other years.

    Since I had so many reprints of ads in the beginning of my year, I think it might be interesting to add name/value pairs for the date the ad was first printed, the date the ad was last printed, and how many times it ran. I expected that ads would be reprinted with each edition of the newspaper, but the one ad that was reprinted most in 1845 was reprinted sporadically. The newspaper was pretty consistently four pages long, so I wonder if some ads were not included in certain editions if they did not fit. For the ones that were only two pages long, I wonder if the other pages of the newspaper were lost, or if the newspaper was occasionally shorter. I’m also curious if this is correlated to the events going on in Texas at the time, because starting around the time that the runaway slave ads became scarce, the news about the annexation of Texas became vast. Looking through a year’s worth of newspapers also made me realize how important the context of the ad in the newspaper could be, depending on what one is researching.

    Because I also found a few things that did not fit with the pattern of a runaway slave ad, such as news reports and ads for hire, a few of the sources found could be useful depending on the topic, but I do not know how they would fit into some of the name/value pairs we have discussed or used in homework.

  5. In my ads, I noticed an interesting distinction of intent between ads for slaves and for animals. Though they are advertized in much the same way, with descriptions and rewards, the ads put the blame on the slaves on their own choice, with the active descriptions of “ran away”, whereas animals are listed as “lost”, “strayed” or “stolen”. I find this an interesting acknowledgement of the individuals’ greater freedom of choice and decision making.
    I also noticed a movement in advertisements that would occasionally locate them on the front page. This seemed odd and deliberate for certain issues, and I would be interested to know why those issues were printed like that.