Unlike other print genres, almanacs and the marginalia that readers added to them give us insight into both everyday life in the Romantic century and the organization systems that people used to manage it. Offering information about holidays, university term dates, hours of sunrise and sunset, predictions for the future, and chronologies of historical events, almanacs were vital books for a mass number of readers who recorded their observations and daily activities in their pages. Importantly, almanacs didn’t simply provide a variety of information; they provided diverse readers with the same standard information. Variation certainly existed between different almanacs aimed as specific groups of readers, such as farmers or lawyers; however, the genre as a whole was largely unified (and, indeed unifying) in the basic, utilitarian information it provided users about time, geography, tidal shifts, and history.

While almanacs were crucial reference tools, they were also for many readers repositories of daily observations and records of both public and private life. Studies of life writing from the Early Modern period to the nineteenth century have noted that writing in almanacs was one of the earliest and most common forms of diary keeping. Famous diarists such as William Gladstone and George Washington began their diaries in almanacs, and far more common, of course, were anonymous readers who noted births and deaths, daily activities, and memoranda in their margins (see Figure 1) [Gladstone 1968, 16] [McCarthy 2013, 11]. For example, Adam Smyth has identified the practice of writing personal notes in almanacs as “the most common form of self-accounting in early modern England”  [Smyth 2008, 204].[4] The almanacs’ prevalence and their use as diaries increased in the eighteenth and nineteenth centuries as growing numbers of people were able to purchase, read, and write in their almanacs.

Our own research has recognized that many readers’ notes in almanacs respond directly to the information and forecasts offered by the printed text. These notes often contradict an almanac’s forecasts or record observations about meteorological events that are not predicted by the almanac (see Figure 2).[5] As a result, almanacs are templates for comparing how numerous readers reacted to the same natural and historical events and responded to the same or similar texts.

Such comparisons are assisted by the capabilities of digital tools, and The Almanac Archive is designed not only to showcase the influential printed content of almanacs from the Romantic century, but also to facilitate the large-scale study of reader habits and observations. Although many single copies of eighteenth- and nineteenth-century almanac issues can be found in digital repositories such as Eighteenth Century Collections Online (ECCO), HathiTrust Digital Library, or Internet Archive, The Almanac Archive provides users with distinctive ways of accessing and interacting with this unique genre.[6] The Almanac Archive’s acquisition strategy is motivated by what we might call thickness rather than breadth. Instead of acquiring one digital copy of as many issues of as many titles as possible, our archive aims to present many annotated copies of a single issue. And instead of prioritizing transcriptions of the printed content of almanacs produced with OCR or human transcriptions, our archive focuses on the accessibility of textual and material features of these texts. To this end, we are using carefully tagged page images rather than textual transcriptions as the primary organizational units of the database.[7] When completed, our resource will permit users to search for a variety of fields related to both the material and the textual features of almanacs from the binding of a particular copy of an almanac and its original cost to the notes, marks, and drawings made by eighteenth- and nineteenth-century readers. Users could, for example, search for any annotations related to the weather made in the month of January 1801. The search results may well turn up multiple copies of a single issue of a given almanac, enabling users to compare how different contemporary readers responded to the same information and the same text. Beyond simply reproducing almanacs for readers in the digital age, then, The Almanac Archive will expose various and unique ways of organizing time and recording personal history and will facilitate comparisons of reader observations. Moreover, as the following section will discuss, these foci make visible the layered interactions between text and Romantic-era reader that are cognate with those between digital text and user.

Almanacs are interfaces that embody ideologies about personal and historical time as well as individual and national experience.[8] Drucker’s work on theoretical approaches to interface design applies to our approach to representing this genre in digital form, for we are not only interested in the almanacs as interfaces but in reflecting on the practice of making the interface through which users of The Almanac Archive will engage with its content. In particular, Drucker argues for interface design that makes evident the role of performativity in systems of organization. She explains: “Multiple imaging modes that create palimpsestic or parallax views of objects make it more difficult to imagine reading as an act of recovering truth, and render the interpretative act itself more visible”  [Drucker 2013, 39]. The Almanac Archive avoids conveying a uniform idea of the Romantic-period reader by representing how different readers approached, organized, and interpreted both time and their books in unique ways. While the archive as a whole — its presentation of published calendars and readers’ annotations about their lives and schedules — may reveal certain patterns of thinking about time, individual copies with readers’ annotations speak to the relational and interpretative nature of reading and annotation. The resource, then, offers the type of palimpsestic representation that Drucker has so convincingly praised. Thus, The Almanac Archive is not primarily about discovering truths about almanacs or annotation practices in the Romantic century, but instead it is about conveying the numerous ways that readers adapted and subverted, employed and rejected the structures of truth and temporality that individual almanacs seemed to promote. Moreover, in allowing digital users to access and analyze marginalia from different readers, The Almanac Archive will expose that historical users of the almanac interface did not, in fact, view their own “reading as an act of recovering truth” but rather as a layered act of reading and writing that overlaid printed facts with handwritten (and sometimes contradictory) observations [Drucker 2013, 39].

Exploring readers’ annotations of and engagements with almanacs speaks to recent trends in the study of reading history. H.J. Jackson’s Marginalia: Readers Writing in Books (2002) and Romantic Readers: The Evidence of Marginalia (2005) are notable examples of this growing interest in readers’ annotations and their potential value for tracking the history of reading. Even more recently, several digital projects have sought to represent the history of reading and marginalia through the creation of databases. For instance, Annotated Books Online (ABO) is “a digital archive of early modern annotated books” that gives users “full open access to these unique [annotated] copies, focusing on the first three centuries of print”  [Annotated Books Online 2014]. Spearheaded by the Universiteit Utrecht, ABO includes a variety of materials from more than ten libraries and has several major partners, including The Centre for Editing Lives and Letters (CELL) at University College London, the Netherlands Organization for Scientific Research, and the University of York. Like ABO, The Archeology of Reading in Early Modern Europe is large in the scope of its primary materials and its partners, which include Johns Hopkins University, the Princeton University Library, and CELL. The project “will explore historical reading practices through the lens of manuscript annotations preserved in early printed books” and has been awarded a $488,000 development grant from the Mellon Foundation [Shields 2014]. On a more grassroots level, Book Traces, sponsored by NINES and the University of Virginia under the leadership of Andrew Stauffer, “is a crowd-sourced web project aimed at identifying unique copies of nineteenth- and early twentieth-century books on library shelves”  [Book Traces 2014] (emphasis in original). Rather than focus on readers per se, Book Traces acts as an argument for preserving unique copies of nineteenth- and twentieth-century books that are at risk of being lost or discarded after the content of a clean copy of the same text has been digitized.[9]

While each of these resources is unique, they are alike in the relative breadth of their focus. By selecting a specific genre of books published within a specific period and geographical location, we aim to create a much more focused dataset, which highlights the importance not just of annotated copies but of seemingly duplicate copies of almanacs that have been annotated. The originality of the resource is that it seeks to present and catalogue as many copies of the “same” almanac as possible to facilitate comparison between readers and different modes of reading. In presenting multiple “duplicate” copies of almanacs and showing the diverse uses to which users have put them, The Almanac Archive presents digitization as an argument for physical preservation, and the project challenges the idea that material books are “in a kind of competition with their own [digital] surrogates”  [Stauffer 2012, 336]. The Almanac Archive uses digitization to draw attention to the importance of retaining duplicate copies that, in actuality, are quite different.

Moreover, The Almanac Archive’s focus on 1750-1850 is especially relevant to duplicate copies, since the rise of the machine-press period in the early nineteenth century made mass producing printed texts a real possibility for the first time. Indeed, the Romantic century encompasses a vital and volatile time in the material production of printed knowledge, yet the media transition from the hand-press era to the machine-press era is one that is all too often simplified and has led to unfounded, persistent assumptions about the “sameness” of books from the machine-press period. Many scholars working across disciplines and temporal periods acknowledge the crucial changes in material knowledge production that marked the transition from the hand-press to the machine-press period at the end of the eighteenth century. Laid paper shifted to wove paper and, soon thereafter, to machine-made paper. The wooden press that had remained substantially the same since Gutenberg gave way to the iron-hand press in the early nineteenth century followed by an assortment of other (steam-powered) presses that produced printed materials in unprecedented quantities. Yet it wasn’t just the scale of print that grew; the uniformity of the printed materials themselves also increased. The prominence of edition binding, for instance, expanded rapidly from the 1820s onward, so that the standardization of the covers of printed books aligned with the perceived uniformity of their printed content.[10] Such a narrative, of course, obscures the granularity of technological changes while representing the production of material artifacts as a flat teleology of technological advancement occurring at the expense of text’s physical uniqueness or “aura,” to borrow Walter Benjamin’s term [Benjamin 2001].

In fact, the transition between the hand-press and machine-press period was gradual and the ability and desire to create uniform, duplicate copies has been overestimated. As bibliographers have long realized, the variety and uniqueness associated with early modern books is also “true of books printed after the hand-press period”  [Stauffer forthcoming]. One way to quantify this variation is to examine not only “duplicate” copies of individual texts or editions (which Stauffer has done with ten copies of the 1902 edition of the poet James Whitcomb Riley’s An Old Sweetheart of Mine, each containing significant bibliographical variants [Stauffer forthcoming]) but also to trace a largely-standard genre such as the almanac through this period of media transition. Comparing duplicate almanacs as well as examining runs of particular almanac titles across decades challenges the notion of sameness and duplication in historical forms of printing and reading practices.

Yet, even while we advocate for digitizing and preserving duplicates, we still do struggle with the logistics of digitally representing them. The question we return to again and again is how much physical variation or annotation makes a copy unique for the purposes of our project? A single word written in the margins? A doodle? A mark? And how do we describe indecipherable or amorphous marks in our database? In posing these questions, we recognize the necessity of balancing our resource’s comprehensiveness with its usability. We are reluctant to flood users with redundant search results containing duplicates with only the most minor or indecipherable annotations, but we also want to create a resource that users with research questions very different from our own can employ to their advantage. The challenge, which continually returns us to the most interesting and fundamental of digital humanities questions, is how to create a database structure that categorizes and tags information in a way that permits users to navigate and search effectively, while also minimizing interpretive decisions on the backend that foreclose client-side interpretative possibilities. For example, we initially considered dismissing some of the less intelligible marks we encountered in almanacs — short horizontal lines penciled next to given dates. However, even these small marks hold important research potential. Maureen Perkins’s study of nineteenth-century almanacs notes that women may have used almanacs discretely to keep track of their menstrual cycles, for example, and such marks may constitute interesting and valuable data for researchers interested in women’s studies or the history of sexuality [Perkins 1996, 44]. The variety of uses that readers brought to their almanacs — and which we are trying to illuminate through our own resource — reminds us that digital resources risk precluding user interpretations through the decisions their designers make while structuring data. To mitigate these lost opportunities as much as possible, we aim for transparency regarding the organizational principles structuring The Almanac Archive. An article such as this one and our decision to make available the archive’s Metadata Application Profile (currently in progress) embodies our desire to be unabashedly open about the project’s development. Our goal is to make clear the choices that have gone into the archive’s design from its initial inception to its actual encoding by emphasizing on the frontend the inherently interpretative choices that have gone into designing the backend.

Our decision to use page images rather than printed text as the primary organizational units of the database is one means by which we hope to widen our resource’s versatility, and it also stems from our questions about current trends that privilege printed textual content as the primary unit of analysis. Privileging print risks de-materializing books through digitization because it substitutes an object’s original organizational structures for a “bag of words” approach that, while useful for text-mining, has theoretical limitations. Focusing on the unit of the page in The Almanac Archive’s design emphasizes and attempts to preserve the importance of the page as a conceptual unit that shaped almanacs’ historical uses in significant ways [Mak 2011]. Prioritizing the page and supplementing it with metadata, The Almanac Archive attempts to limit the appropriative, flattening elements that Hillel Schwartz associates with copying technologies that take “without homage” and obscure “the historical steps that gave rise” to the original and, in turn, its original uses [Schwartz 2014, 191].

Theorizing The Almanac Archive has therefore led us to think critically about prominent ways of categorizing digital objects. In “Digitizing Latin Incunabula: Challenges, Methods, and Possibilities,” Jeffrey A. Rydberg-Cox describes five different categories of digitization: Image Books, Image Books with Minimal Structural Data, Image Front Transcriptions, Carefully Edited and Tagged Transcriptions, and Scholarly and Critical Editions [Rydberg-Cox 2009, 8]. The Almanac Archive fits into the category of Image Books with Structural Data, yet we resist the connotations of Rydberg-Cox’s term “minimal,” for the metadata attached to each image and almanac will allow refined searching by individual categories, such as publisher, place of publication, library collection, type of almanac (i.e. astrological or agricultural), and type of annotation (i.e. textual or pictorial). Thus, we challenge the idea that there is a hierarchy of digitization involving various degrees of access to textual content. Instead we conceive ways of digitizing that prioritize the bibliographic unit of the page. Because the project recognizes and highlights that almanacs were physical interfaces that users interacted with, we want to avoid creating a digital interface that privileges either printed text or manuscript annotations, but instead always displays the two side by side, in relation to each other. John Bradley’s work on digital annotation has emphasized how annotated texts involve “two rather different applications [that] must co-exist”  [Bradley 2012, 12]. While we agree that the different agents, intentions, and technologies of the publishers and the annotators must be recognized, our archive seeks to emphasize the codependence of these two applications. In this respect, our approach prefers thinking of these applications as “overlapping hierarchies,” as Joanna Drucker does, thereby recognizing the competing ways of valuing different linguistic and bibliographic components of a digitized object [Drucker 2007, 182].

The unique research potential of our archive is a function of the fascinating corpus and little-studied genre we have chosen and of our commitment to making both bibliographic and linguistic features of the almanacs accessible to users. By valuing equally annotations, printed text, as well as key metadata that will be included in the archive, we resist privileging one type of information over another. For our own research we will use the resource to track how eighteenth- and nineteenth-century readers related to time and historical events, and we will therefore focus primarily on the annotations found in the almanacs to compare reader responses to time. Yet annotations also offer other avenues for scholarly research. For readers unable to afford both an almanac and a dedicated pocket-book diary, these volumes functioned as diaries. Because they were so inexpensive, often costing between just one and three shillings, almanacs were affordable to people who were unable to purchase other printed matter. Given the characteristics and uses of the almanac in this period, the potential applications for this project are diverse. Historians of weather, for instance, might track weather patterns or climate change since users frequently noted weather anomalies. Thus, material from The Almanac Archive might support other digital resources such as Old Weather, which invites users to transcribe weather-related entries found in ships’ logs. Some almanacs include annotations about payments to employees and other financial records, indicating that they may be untapped resources for information about labour and economic history. Readers’ notes in farming almanacs provide insight into historical agricultural practices.

While in some instances information about an almanac’s annotator remains unknown, in others inscriptions of ownership or records of provenance provide key information about annotators such as gender, occupation, geographical location, and socioeconomic status. Aggregating these entries and this information will, as the corpus expands, offer a new tool for historians, particularly those interested in questions of everyday life and gender in Britain.

As compelling as the almanacs’ annotations and annotators are, we are designing The Almanac Archive to foster other potential areas of research. For example, the database will contain a metadata field for the prices of almanacs. This information about price combined with information about readers’ marginalia could help researchers make more concrete arguments about annotation practices related to cost of books. The metadata will also include information about provenance and how copies of almanacs are organized in library collections. We have, for instance, noted that some users bound a series of almanacs together. As our corpus of almanacs grows, it will be possible to draw conclusions about not only the culture of annotating individual volumes but also larger cultural value systems that impacted how groups of almanacs were preserved and organized. As Thomas R. Adams and Nicolas Barker have suggested, the survival and durability of different kinds of texts is an important variable in any model of book history [Adams and Barker 2001].

Of course, just as eighteenth- and nineteenth-century readers used the interfaces of their almanacs in surprising ways, we anticipate that once the resource becomes live scholars and teachers will use it in similarly original ways. The flexibility and the variety of our archive’s applications, then, seem to embrace in a digital, scholarly form the original aspects of the historical artifacts that will comprise the archive itself. Ultimately, our archive’s unique research potential lies in the light it sheds on the diverse and unexpected ways users respond to structured data, whether in eighteenth-century printed texts or in contemporary digital resources.