Source Criticism
The widespread use of born-digital assets necessitates the development of source criticisms. For instance, emails are quintessential born-digital documents that constitute a valuable repository of historical knowledge. However, in studying emails, historians must be cognizant of particular features of the email system – what was “marked as read”, what was ignored in the inbox, and, indirectly, the user interface and algorithmic rules of the email system (Owens & Padilla, 2021).
Fingerprints, dirt on pages indicate what has been read. For examination of these physical, material aspects of a source, only digitized materials of very high quality are suitable for historical analysis. Conversely, the obfuscation of some material or other aspects during digitization, as well as the digitization process itself, must be considered. The overarching question that emerges from these archives is then whether “digital surrogates” for the original, analog material are adequate or sufficient.
Digital instantiations of archives enable simultaneously searching across a vast number of sources. Such capabilities greatly expand the reach of the methods employed in historical and humanities scholarship. A typical example is the full text search, which has become an indispensable research tool. Although these tools greatly facilitate research, they raise new questions concerning the interpretation of these sources, and reasoning from and about them. Formerly, searches and research were contextualized. It was not possible to explore an entire repository or archive. Research questions were directed. With new, algorithmic search tools, historians can perform less structured searches, looking instead for interesting things. However, once found, these interesting aspects must still be contextualized, and the sources and motivations for what was found must be investigated (Owens & Padilla, 2021).
There is another consideration that is especially relevant for historical research. Because digitized sources are binary values encoded on some storage medium and are essentially representations and surrogates that are viewed on a 2D computer screen, some of the material, analog aspects of the original source are not available. Some scholars have criticized the “screen essentialism” for the reliance on how materials appear on computer monitor. However, in digital files, file systems, and databases, the reader may be viewing additional encoded information. Metadata and other information, such as email headers for instance, can provide supplementary information beyond what can be displayed on a screen. As an example, in a born-digital document created with word processing software, creation information is available, such as the initial date and time of the document. In addition, edits made to a document over time are stored in the data associated with that document, and can subsequently be accessed, allowing the researcher to examine how the content has been modified. However, to productively utilize this additional information, the researcher must have some knowledge of the software and technology used to create, write, and edit the document, as well as a basic understanding of the software and technology used by the researcher. Furthermore, on digital storage media such as floppy disks, the predecessors to CD-ROMs, “deleted” data are not technically deleted, but are instead marked by the operating system as being available for overwriting. In other words, unless the “deleted data” has been overwritten with new data, it is available for inspection, usually through specialized programs that read binary codes from media (Owens & Padilla, 2021). Knowledge of digital tools can therefore open many exploratory possibilities for technologically savvy researchers.
Web pages on web archives are rendered differently with present-day web browsers than they were originally. Limitations in screen resolution, processor speed, and memory dictated what could be done on web pages in the past. Therefore, researchers viewing these archived web pages in the present are not experiencing the sites in the same way as when the page was created.
Email messages from cell phones – “Sent on my … device” implies constraints, such as the size of the input interface, thereby providing clues as to the presence of typographical or spelling errors, or compression.
Digital pictures presented on web sites are usually not the original born-digital photograph. They have photograph transferred to different photo viewers and may have undergone subtle transformations through image processing and filtering. However, as with all digitized materials, metadata embedded into the images – including the software used to edit them – may provide additional useful information to researchers (Owens & Padilla, 2021).