In Which Digital Surrogates Must Be Used as Complements to, Rather than Replacements for, Original Sources

In recent weeks I have questioned some of the decisions made by the printers of the New-Hampshire Gazette in the early months of 1766. Several of the advertisements the guest curators have selected featured rather unconventional formats, sometimes as a line or two running across the bottom of a page and other times entire advertisements tilted counterclockwise ninety degrees. I discovered other irregularities as I consulted Early American Newspapers to see these advertisements in the context of the entire printed page. What were the printers thinking?!

I’ve suggested a couple of theories. For instance, I hypothesized that some of the unusual layout might have been the result of the printers attempting to create front and back pages with equal coverage of text. I also questioned whether some of the decisions about the layout might have been innovative experiments to draw readers’ attention to the advertising. In both instances, I promised to consult the originals, rather than the digital surrogates, to see if that would answer any questions.

It turns out that examining original issues of the New-Hampshire Gazette helped me to solve the mystery! This is an instance that demonstrates that digital surrogates cannot replace original sources. Instead, the two complement each other, allowing scholars to ask new questions and find answers that would not have been possible (or, at the very least, much less likely) when relying on only one or the other.

Here’s what was happening with the New-Hampshire Gazette in 1766. Like other newspapers of the era, this one was published on a broadsheet that was folded in half to create an issue with four pages. In this case, each page measured 9½ inches wide by 14¾ inches high. Each page featured three columns whose width each measured 2¾ inches. Let’s call that a “standard” format for the purposes of this examination.

For the February 7 issue, however, the printers switched to smaller broadsheets. I imagine that the larger ones were in short supply or not available. That issue, as well as the next two (February 14 and 21) still included four pages, but each page was 8 inches wide by 13 inches high rather than 9½ inches x 14¾ inches. Instead of three columns measuring 2¾ inches wide each, the new size accommodated two columns that were 3½ inches wide (in most cases, but I will soon describe some of the very interesting exceptions). Let’s consider these new parameters a “temporary” format for the purposes of this examination.

Starting with the February 28 issue, the newspaper returned to the larger size broadsheet, but not to any sort of “standard” format. That and the following issue, March 7, featured advertising layouts that were even more visually jolting. With the March 14 issue the New-Hampshire Gazette returned to publishing “standard” format issues of four pages with three columns measuring 2¾ inches wide each.

Although Early American Newspapers revealed to me the difference in the number of columns, it did not offer any metadata explaining the transition. The difference was immediately apparent, however, when I viewed the originals at the American Antiquarian Society. I spent quite a bit of time with a ruler measuring the pages and the columns for each issue from January 31 through March 14.

This particular kind of metadata is currently lacking in Early American Newspapers. That is an observation rather than a complaint. I realize that it would be prohibitively expensive to develop this particular metadata: all the measuring, recording, coding, and other tasks required to link this data to the digitized images that substitute for the original newspapers. As somebody who regularly uses digitized newspapers I would much prefer that the partnership of the American Antiquarian Society and Readex continue to produce digital images of as many newspapers as possible, allowing for even more expansive research by scholars working on a variety of projects and members of the general public asking a variety of questions. I understand that the trade off for access to a greater number of sources will often be less metadata for most of them. Still, I feel that the strategy that has been pursued here has been appropriate, as long as those using the digital surrogates are aware that the digital images cannot reveal everything that the original sources do.

This raises important questions about what we miss when we rely exclusively on digital sources. I first began thinking about this issue as a result of Leon Jackson’s presentation at the Digital Antiquarian Conference last summer: “Historical Haptics: Digital and Print Cultures in the Nineteenth Century.” (Check out the slides from his presentation.) He pointed out that every issue of a newspaper appears to “be” the same size when consulting digital sources: the size of the screen. That was certainly my experience with the New-Hampshire Gazette. The digital surrogates hid some important information – the size of the page – from me.


Let’s have a look at some of the unusual layout decisions the printers made when they temporarily used smaller sheets for their newspaper.

NH Gaz 2:7:1766 Third Page
Third page of the New-Hampshire Gazette (February 7, 1766).

Consider the third page of the February 7 issue. The first two pages featured the “temporary” format: two columns that were each 3½ inches wide. The third page, however, had two columns supplemented by a third column of four advertisements each rotated ninety degrees counterclockwise. It turns out that each column measures 2¾ inches wide, as do each of the four advertisements in the extra column with perpendicular text. Each of those advertisements previously ran in earlier issues. The printers likely still had the advertisements in the forms. Rather than reset the type for the new format of the newspaper they saved time by reusing the work they had already done (and keep in mind that they would have reused it all the same if they had access to the larger broadsheets that week).

This issue also included advertisements featured here previously, both from Jonathan Moulton. On the first and fourth pages they ran for two lines across both columns at the bottom of the page. Why? Perhaps this was a means of squeezing in a little more advertising. Perhaps it was because Moulton offered a March 1 deadline for purchasing cows, making his notice time sensitive, prompting the printers to insert it even after the rest of the issue had been set for printing.

NH Gaz 2:14:1766 Second Page
Second page of the New-Hampshire Gazette (February 14, 1766).

The second page of the February 14 issue included other curious decisions about the layout. At first glance it appears that the printers followed the same course as they did for the third page for February 7. On closer examination, however, it turns out that the first column measure 3½ inches wide while the second is only 2¾ inches wide. The four advertisements rotated to form the third column all measure 2¾ inches wide.

With one exception, the first column features either new items or advertisements repeated from the previous issue and set with a 3½-inch width at that time. The exception is a “new” advertisement from Moulton which combines the two that appeared on the first and last pages of the previous issue into a single advertisement (which was then repeated in subsequent weeks).

All of the advertisements in the second column, measuring 2¾ inches wide, were repeated from issues printed on larger broadsheets on January 31 or earlier. Again, the printers reused type already set in the forms.

Careful attention to the perpendicular column reveals that all four advertisements measure 2¾ inches wide. Each was repeated from type set for the “standard” format. However, one of them underwent minor alterations to make it fit into the space available. The font size for “EDWARD EMERSON” (formerly “Edward Emerson”) was reduced and the advertisement was truncated, removing “at the lowest Rates” and shaving off a line of text as a result.

The February 21 issue, the final one printed on smaller broadsheets and featuring the “temporary” format, had a very regular appearance compared to the previous two. All four pages had two columns that measured 3½ inches wide each, giving the issue a rather clean appearance compared to those that came immediately before it (and, as we will see, those that came after). The only exception was a one-line advertisement that ran across both columns at the bottom of the second page.

NH Gaz 2:28:1766 Third Page
Third Page of New-Hampshire Gazette (February 28, 1766).

Printed once again on the larger broadsheets, the February 28 issue returned to the “standard” format of three columns, each measuring 2¾ inches wide, on every page except the third one. That page had what was becoming a familiar layout: two columns accompanied by a narrow column of advertisements rotated ninety degrees counterclockwise. Unlike either of the previous examples, however, the first and second columns measured 3½ inches wide each. All of the content was either new or repeated from issues for which type had been set to accommodate the smaller broadsheets. Again, the printers efficiently used type already in forms.

That final rotated column is definitely the most curious. The printers certainly attempted to maximize content. It featured three items that measured 2¾ inches wide: a news item from London, a new probate notice, and Edward Emerson’s advertisement (restored to its original format). The printers also inserted an advertisement that was 3½ inches in width that had originally appeared in the February 14 issue. It was too long to be wedged in along the right edge of the broadsheet so the printers removed the final four lines and inserted them as a narrow (1 inch wide) addendum next to the rest of the advertisement, giving the upper right corner of the page a rather strange appearance.

NH Gaz 3:7:1766 Third Page
Third page of the New-Hampshire Gazette (March 7, 1766).

The printers got even more creative/efficient with the layout of the third page of the March 7 issue. (All the other pages of that issue returned to the “standard” layout.) The upper quarter of the page featured two columns measuring 3½ inches wide each and a now-familiar third column with the text rotated. All of the material in the first two columns was either new or repeated from an issue that had been set intentionally with column widths of 3½ inches.

In order to maximize the number of advertisements they could squeeze into that rotated column, the printers chose two that were 3½ inches wide and a third that was 2¾ inches wide. This allowed them to fill the page in the upper right corner. As with Edward Emerson’s advertisement previously, the font size for the first line of John Wheitfield’s advertisement was reduced in order to make it fit on the page. Similarly, the first line of George Jaffrey’s notice about excise taxes on liquor (“PROVINCE OF NEW HAMPSHIRE”) was removed in the interest of space. The probate notice was left intact from the previous issue.

The lower portion of the page included three columns, each 2¾ inches wide. The printers were continuing to transition back to the “standard” format and setting new type accordingly now that they had access to the larger broadsheets once again. The first two columns included a fresh news item from London and a new advertisement for a “Milch COW.” The third column included slightly updated advertisement by shopkeeper Joseph Bass. To save space, the first line (“Just imported from LONDON”) and a nota bene at the end were removed.

All told, the third page of the March 7 issue is the most visually striking. It also most effectively demonstrates the printers’ creativity and efficiency in printing on the paper that was available to them and conserving their energy and efforts by not completely resetting type for all the advertisements as their paper supply changed and then changed again.

As noted above, the March 14 issue returned to the “standard” format. None of the layout merits further mention, except perhaps acknowledging that the missing lines from Joseph Bass’s advertisement returned now that more space was available. This and other examples in these issues suggest that printers played the primary role in decisions about format even if advertisers called the shots when it came to content.


In the end, I am not able to make the argument that I initially hoped to make about some of the innovative layouts for advertisements in the New-Hampshire Gazette in the early months of 1766. I wish I could argue that the printers were experimenting with the layout as a means of drawing readers’ (and potential customers’) attention to the advertisements, but after consulting the original issues I realize that the evidence just does not bear out that interpretation. What I took for innovation in advertising turned out to be an innovation of necessity and practicality as the printers creatively and efficiently inserted as many advertisements as possible in the wake of printing on smaller broadsheets.

I would not have been able to reach this conclusion had I relied on digitized images of the newspapers alone since that format obscures the size of the newspapers and the changes in the paper supply. As much as possible, original sources and digital surrogates really must be consulted in combination as complements to each other.

In Which One Digital Archive Nearly Replicates One Existing Archive

In recent weeks I’ve spilled a fair amount of ink considering how both methodology and access shape the Adverts 250 Project. I’ve demonstrated three different levels of access to newspapers printed in 1766 included in Early American Newspapers: 14 via my college’s library, 15 via the Boston Public Library’s electronic resources (including the extremely significant addition of the Pennsylvania Gazette) and 21 via the digital resources available in the reading room at the American Antiquarian Society (which has access to all the titles in Early American Newspapers because the AAS and Readex are partners in the endeavor).

I’ve also demonstrated how my methodology for selecting advertisements (each must come from that date or a newspaper published most immediately before that date in cases of no newspapers printed on a particular date) has caused certain newspapers to receive disproportionate coverage due to most newspapers being published at the beginning of the week and relatively few at the end.

Feb 19 - Masthead for New-York Gazette 2:18:1766
Masthead for an Extraordinary (“Extra”) to the New-York Gazette (February 18, 1766).

I’ve made promises that when my Public History students’ tenure as guest curators comes to an end that I will resort to the resources available in the AAS’s reading room as a means of featuring a greater number of publications and achieving more extended geographic reach.


However unintentionally, I may have implied that accessing Early American Newspapers at the American Antiquarian Society means that I am working from a complete archive of publications from 1766. There are several reasons why this assumption is not completely accurate. The project will be migrating toward the best possible digital access, but that is not the same as complete access to every newspaper in an archive. Today I’d like to examine how closely the most extensive access to Early American Newspapers mirrors what is available in the stacks at the AAS.

Recall that I previously identified these newspapers printed at some point in 1766 (from Edward Connery Lathem’s Chronological Tables of American Newspapers, 1690-1820) that are not available via even the AAS’s most extensive access to Early American Newspapers.

New Hampshire

  • Portsmouth Mercury (last known September 29)

New York

  • New-York Gazette or Weekly Post-Boy


  • [Germantown] Wahre und Wahrscheinliche Begebenheiten (only known February 24)

North Carolina

  • [Wilmington] North-Carolina Gazette (last known February 26)

South Carolina

  • [Charleston] South-Carolina and American General Gazette
  • [Charleston] South-Carolina Gazette (suspended starting October 31, 1765; resumed June 2, 1766)
  • [Charleston] South-Carolina Gazette and Country Journal
Feb 19 - Masthead for Connecticut Courant 2:17:1766
Masthead for the Connecticut Courant (February 17, 1766).

I consulted the AAS’s online catalog and, especially, Clarence to find out if the AAS collections included these newspapers. (Clarence – named for Clarence Brigham, librarian (1908-1930) and director (1930-1959) of the AAS and author of the two-volume History and Bibliography of American Newspapers, 1690-1820 – is a database that indicates which specific issues of early American newspapers are in the AAS collections, replicating and updating portions of Brigham’s monumental bibliography.) Here’s what I discovered:


  • Portsmouth Mercury: scattered issues from 1766.
  • New-York Gazette or Weekly Post-Boy: weekly issues and occasional supplements through October 9.
  • Wahre und Wahrscheinliche Begebenheiten: photostat copy of only known issue (as indicated in the catalog record; Clarence does not specify this detail).
  • North-Carolina Gazette: AAS does not possess any issues.
  • South-Carolina and American General Gazette: scattered issues from 1768 through 1778, but none from 1766.
  • South-Carolina Gazette: one damaged issue from 1766 along with scattered issues from 1737, 1740, 1760, 1763, 1767, 1768, 1770, 1772, and 1774.
  • South-Carolina Gazette and Country Journal: scattered issues from 1766 as well as scattered issues from 1768 through 1775.

Based on these findings, it appears that digital access to newspapers printed in 1766 via Early American Newspapers very nearly replicates the holdings of the American Antiquarian Society. For the most part, the AAS possesses only scattered issues of the titles not included in Early American Newspapers. This is a major achievement that allows researchers to view the contents of these publications while also preserving the originals.

For the purposes of this project, a digital archive that nearly completely replicates the newspaper holdings of the AAS (at least, those printed in 1766) also streamlines the research process. Having compiled a calendar of which newspapers were printed on which dates in 1766, I can quickly scan the relevant issues when selecting an advertisement to feature on any given date. I imagine that the reading room staff at the AAS also appreciates that I am not repeatedly requesting large bound volumes of eighteenth-century newspapers that they then have to page, process, deliver to me, and later return to their designated places in the closed stacks.

On the other hand, for other sorts of projects, the research process goes much more smoothly and efficiently when I can quickly – but carefully – flip through the pages of a bound volume of newspapers, scanning for particular content. To preserve the originals, scrolling through microfilm copies serves the same purpose. When it’s necessary to examine a large number of issues published sequentially, digital access via Early American Newspapers can be slow and cumbersome by comparison. That’s not a criticism but rather recognition that digital surrogates are not always the best format for conducting research. (On the flip side, the ability to do keyword searches in Early American Newspapers can streamline the research process significantly. I’ll write more about the virtues and imperfections of keyword searching digitized newspapers some other time.)

Feb 19 - Masthead for Boston Evening-Post 2:17:1766
Masthead for the Boston Evening-Post (February 17, 1766).

I noted above that in the coming months this project will migrate to the best possible digital access, but that is not the same as complete access to every newspaper in an archive (although in this case it is really close). Next week I will consider the difference between access to every newspaper in an archive and access to every newspaper printed. Once again, these distinctions may seem merely academic at first glance, but I continue to maintain that researchers must be aware of the scope and limitations of their resources and we have an obligation to others who read our work to share that information.


As I consider these issues, I keep returning to two of the main arguments presented by Kenneth Carpenter (Harvard Libraries, retired) and Michael Winship (English, University of Texas – Austin) in their keynote address at the Digital Antiquarian Conference last May: (1) digital sources should be consulted as complements to, rather than replacements for, original sources and (2) be conscious of the metadata that provides the foundation for digitized sources so you know how closely digital surrogates replicate original documents.

In Which Methodology, as well as Access, Significantly Shapes the Project

Last week I demonstrated that different institutions have varying levels of access to the titles included in Early American Newspapers, arguing that this shaped the scope of the project. The access from the reading room at the American Antiquarian Society includes 21 newspapers published in 1766 with broad geographic reach, while accessing Early American Newspapers via my college’s library allows me to examine only 14 newspapers. Those titles are confined mostly, but not exclusively, to New England and New York.

In some ways the effects might be minimized, especially if we take into account T.H. Breen’s argument about the standardization of consumer culture throughout the colonies in the eighteenth century. Colonial consumers, he maintains, purchased imported goods that were increasingly uniform from port city to port city, region to region, in the decades leading up to the American Revolution. Indeed, Breen claims that having similar experiences in the marketplace and speaking a common language of consumption facilitated Americans’ ability to speak to each other about political matters, especially as they imbued consumption with political valences in the wake of the Stamp Act and other measures enacted by Parliament.

That being said, I would still prefer to demonstrate that advertisements from the Chesapeake and the Lower South marketed the same or similar goods and deployed the same or similar appeals as the commercial notices printed in newspapers from New England and the Middle Atlantic. I would like to be able to show – visually and through commentary – rather than merely tell. In addition, even if colonists did have access to increasingly standardized goods and services throughout the colonies that does not necessarily mean that regional differences did not also emerge, especially in terms of marketing. Did advertisers throughout the colonies make similar appeals as they marketed the same assortment of goods?

These are some questions I wanted to raise last week, but the bibliographic details were already so extensive that I held them in reserve. This week I would like to raise some other questions about the contours of this project, but these are based on my methodology rather than access to Early American Newspapers.

Recall that whenever possible I select an advertisement published on that date 250 years ago. When no newspapers were published on that date (or, at least, none that I can access!) I resort to a newspaper printed as close to that date as possible (but always previously printed: it must have been in the hands of colonists somewhere). In addition, I consult newspapers from as many different cities, colonies, and regions as possible. On some days I have multiple options. To help illustrate this, guest curator Kathryn J. Severance and I worked out this census of newspapers that we can access via the Boston Public Library’s electronic resources for the current week, Sunday, February 7 through Saturday, February 13.

February 6, 1766

  • Boston News-Letter
  • Pennsylvania Gazette
  • These newspapers fall outside of this week. Consult them only if there were no acceptable adverts in the newspapers published on February 7.

February 7, 1766

  • Connecticut Gazette
  • New-Hampshire Gazette

February 8, 1766

  • No newspapers were published on February 8. Use any newspaper published on February 7 or, if necessary, February 6.

February 9, 1766

  • No newspapers were published on February 9. Use any newspaper published on February 7 or, if necessary, February 6.

February 10, 1766

  • Connecticut Courant
  • Boston Evening-Post
  • Boston Gazette
  • Boston Post-Boy
  • New-York Gazette
  • New-York Mercury
  • Newport Mercury
  • Wochentliche Philadelphische Staatsbote

February 11, 1766

  • No newspapers were published on February 11. Use any newspaper published on February 10.

February 12, 1766

  • No newspapers were published on February 12. Use any newspaper published on February 10.

February 13, 1766

  • Boston News-Letter
  • Pennsylvania Gazette

Notice that no newspapers were published on February 8 and 9, 1766. For those dates the methodology dictates selecting advertisements from either the Connecticut Gazette or the New-Hampshire Gazette. I know from experience that both of those can be rather slim pickings when it comes to advertisements for consumer goods and services. The Connecticut Gazette often did not feature any, while the New-Hampshire Gazette tended to reprint the same advertisements for multiple weeks.

This census helps to illustrate the somewhat surprising origins of the first 30 advertisements featured in 2016 (before my Public History students assumed their guest curator responsibilities). I included this list last week, but did not have sufficient space to evaluate it.

  • 7 advertisements: Massachusetts Gazette (Boston)
  • 7 advertisements: New-Hampshire Gazette (Portsmouth)
  • 4 advertisements: Newport Mercury
  • 3 advertisements: New-York Gazette
  • 3 advertisements: New-London Gazette
  • 2 advertisements: Connecticut Courant (Hartford)
  • 1 advertisement: Boston Evening-Post
  • 1 advertisement: Boston Gazette
  • 1 advertisement: Boston Post-Boy
  • 1 advertisement: New-York Mercury

In seeking to be current, to provide “the freshest advices foreign and domestick,” my methodology gives disproportionate attention to the New-Hampshire Gazette, a relatively minor newspaper from a relatively small town. (I know, I know: harsh words for what now bills itself as “The Nation’s Oldest Newspaper,” having been in continuous publication in one form or another since 1756.) The New-Hampshire Gazette did not publish nearly as many advertisements for consumer goods and services as its counterparts in Boston, New York, and Philadelphia. That the New-Hampshire Gazette has been featured so prominently is a consequence of selecting advertisements from the most recently published newspaper on any given date.

The New-Hampshire Gazette and the New-London Gazette, which included even fewer advertisements, were printed on Fridays in 1766. No newspapers were printed on Saturdays or Sundays. As a result, my methodology prescribes that I select advertisements from these two publications three days of the week. Sometimes neither featured enough advertisements to make this possible, forcing me to go back to the Massachusetts Gazette, which also helps to explain why so many advertisements featured here derive from its pages. (Keep in mind that I used accessed Early American Newspapers via my college throughout January, which meant that the Pennsylvania Gazette was not an option. Since my students are using the Boston Public Library’s electronic resources to access Early American Newspapers they have incorporated the Pennsylvania Gazette into this project. When I am once again responsible for selecting the featured advertisement each day I will incorporate an even greater number of publications by accessing Early American Newspapers in the reading room at the American Antiquarian Society.)

I’ll close today’s extended commentary by reiterating that the New-Hampshire Gazette has received disproportionate attention due to the methodology I have developed for pursuing this public history and digital humanities project, not solely due to variations in access to Early American Newspapers. Quite simply, newspapers were printed on the day they were printed. On the other hand, I have developed a much different sampling method for my book project. The methodology I use here would not be appropriate in a manuscript seeking to analyze the development of advertising in eighteenth-century America.