Two rather different answers to the question of how to read a database leap to mind. First, "you don't." This answer insists on a conventional sense of reading and suggests a difference in kind between a personal literacy narrative and a collection of such narratives—especially one held in a digital database. As we saw above, you may browse or search information about the narratives in the database (i.e., the metadata), but you certainly don't read that information in the very important sense of the word in which we read (or view or listen to) individual narratives—the sense in which the guest curators read the narratives featured in their exhibits. Alternatively, we might answer that you read a database of literacy narratives "one item at a time." This answer also focuses on a particular scale or granularity, downplaying—even eliding—the patterns of information that might be revealed by computer-assisted analysis of the entire database. And, truth be told, the current public interface of the DALN—built on an implementation of DSpace hosted by OhioLINK for the Digital Resource Commons of Ohio—provides no tools to users beyond search and browse for exploring the database as a whole.
When considering the process of reading a database, it helps to consider what the results of searches like the one described above reflect. Searches of the DALN database query a subset of the metadata (i.e., descriptive information) associated with each narrative, metadata supplied by contributors using their own language (the DALN does not enforce controlled vocabulary for most metadata fields for which contributors supply information). All but one of the metadata fields represented on the submission form are optional (contributors must supply a title). In addition, searches query the full text of any narratives in text formats (e.g., .txt and .doc) and transcripts of audio and video narratives, when transcripts are available. Finally, the DALN is not indexed in the manner that books have traditionally been indexed; its searches cannot point users to concepts that do not appear explicitly in some text, either in the metadata, narratives, or transcripts.
If searching the DALN doesn't yield satisfactory results, users can browse the DALN by author, title, collection, date of submission, or subject terms supplied by contributors. Again, several caveats apply. Many narratives in the DALN were submitted anonymously. Titles are required, but many contributors enter "Untitled" or "My Literacy Narrative" in the title field. Providing subject terms is optional and not constrained to a controlled vocabulary. Collections created by the DALN team and our contributing partners (see http://people.cohums.ohio-state.edu/selfe2/daln/partners.html) are perhaps the most internally related subsets of the DALN, as they tend to group narratives by institution or a single demographic characteristic, but they do not necessarily contain all of the narratives in the DALN that might belong to that institution or fit that demographic characteristic.
These characteristics (and others) of the DALN reflect deliberate design choices—e.g., lowering the bar for participation by simplifying the submission process, refraining from imposing demographic or descriptive categories on contributors, and building the DALN on the margins of an existing infrastructure managed by institutional sponsors as a permanent cultural archive. In sum, those design choices have helped the DALN grow rapidly and capture unique data about contributors' narratives, but they present some interpretive challenges for users. Without a professionally produced subject index or consistent and complete metadata, users searching or browsing the DALN must consider the narratives themselves and all of the metadata provided with the narratives as posing interpretive problems and potential insights to contributors' conceptions of their literacy stories, and users must think creatively about ways such information might provide useful links to other narratives. With that context in mind, this essay explores the process of reading a database.
Before exploring ways to read the DALN database, however, it is helpful to review some of the interpretive problems posed by reading just one personal literacy narrative from the DALN—as well as the problems posed by reading a small group of narratives about which we posit some relationship, if only to remind ourselves that the interpretive problems posed by the metadata that indeterminately relate items in a database to one another are roughly analogous to the problem of understanding the relationships between individual literacy narratives or groups of narratives and the cultural contexts to which they refer.
The stories collected in the DALN constitute narratives in a very basic sense of the term: "An account of a series of events, facts, etc., given in order and with the establishing of connections between them" (OED). Of course, any given telling of a personal story may not follow strictly chronological order; it may include passages of description, opinion, and reflection as well as accounts of events; and it may not explicitly or clearly establish the connections among events. Nevertheless, only contributions that contain a basic narrative core—some account of personal experience—are accepted for preservation in the DALN. But what can such personal literacy narratives mean? Put another way, what assumptions about narratives in the DALN might transform them into data—information "given or granted; something known or assumed as fact, and made the basis of reasoning or calculation" (OED)—useful for understanding literacy practices and values?
We can say without hesitation that narratives in the DALN provide evidence about the language that some people use when asked, in various circumstances and in various ways, to describe literate behavior—evidence consisting of diction, figurative language, narrative structures, allusions, and so on. Audio and video narratives provide, in addition, data about posture, expression, gesture, tone of voice, and so on that contributors employ when discussing literacy. Audio and video narratives also provide at least some information about the setting in which a given narrative is told and the prompts and interactions that elicit the narrative.
The fact that narratives in the DALN typically draw on memory (though some contributors refer to—and occasionally include—related documents) presents more nuanced interpretive challenges. While it makes sense to assume that contributors to the DALN attempt to speak the truth as they recall and understand it, memories can of course be incomplete, or incoherent, or become indistinct with the passage of time. Recounted memories are no doubt inflected to some degree by all that has happened since the remembered incidents. And stories told from memory are presumably not static versions of events or an individual's interpretation of events; they are one telling, one attempt at understanding. The circumstances of storytelling—context, audience, format, available time, and so on—will all necessarily affect which of the available contents of memory are woven into story, and how.
In short, broad assumptions about the relationship between the narratives in the DALN and the historical events or cultural contexts to which they refer require a number of caveats, even leaving aside the possibility of intentional or unintentional exaggeration, embellishment, or fabrication. Just so, we must read the relationships encoded in the DALN database critically.
How, then, might we further characterize the interpretive problems presented by reading literacy narratives and their related metadata in an archive of over 3,600 literacy narratives? Try substituting "literacy narrative(s)" for "poem(s)" in the following advice from Robert Frost:
The way to read a poem in prose or verse is in the light of all the other poems ever written. We may begin anywhere. We duff into our first. We read that imperfectly (thoroughness with it would be fatal), but the better to read the second. We read the second the better to read the third, the third the better to read the fourth, the fourth the better to read the fifth, the fifth the better to read the first again, or the second if it so happens. For poems are not meant to be read in course any more than they are to be made a study of. (165)
Frost notes, and implicitly suggests the importance of, the arbitrary order in which we encounter texts outside of a deliberately constructed sequence of study, the endless and recursive process of reading texts in light of other texts, and the inconclusiveness of any single reading. Much the same could be argued about reading the literacy narratives in the DALN. Visitors can profitably pick a narrative to read at random, then another and another, reflecting on similarities and differences as they go, making meaning from the literacy narratives in the DALN by charting a unique path through the archive as much as discovering meaning encoded by the contributors.
Of course, Frost's focus on individuals' encounters with texts unencumbered by externally imposed order or purpose inevitably obscures a countervailing truth: texts never present themselves to us absent of contexts that shape and order our understanding of them. If we read a poem in a book containing other texts, the other texts provide one sort of context for understanding the poem. If we encounter a book in a bookstore or library or friend's bookshelves, it will appear in the context of other books, perhaps grouped by subject, genre, or price (e.g., the bargain bin). If the text is ascribed to an author, we are presented with the possibility of reading it as part of the author's oeuvre. If dated, we are presented with the knowledge that the text was published in a chronological sequence of publications, whether or not we choose to read the text in that sequence. Metadata matters.