Metadata is crucial for the digital humanities in several respects, from initially classifying and organizing raw sources to later discovery and analysis by researchers. Put simply, metadata is a common core of descriptive categories that archivists apply to incoming records to facilitate not only their own internal organization of those records, but more importantly to enable easier discovery by subsequent users of an archive. (Greer and Neatrour, 2015) The importance of metadata to an analog database is easy enough to understand, given the infeasibility of manually searching through an often overwhelming amount of physical records; analog metadata discovery tools like card catalogs have long facilitated more efficient discovery of useful sources.

For digital humanists, however, metadata assumes an even greater importance given the central importance of electronic search tools within the field. A relatively small number of digital archival records are fully transcribed and thus discoverable via simple keyword search, particularly when dealing with older records that may be handwritten and thus difficult to digitize via text-scanning processes like Optical Character Recognition (OCR). Furthermore, non-text resources like posters, photographs, artwork, and digital renderings of physical objects are intrinsically incapable of being discovered via any text-based search that is limited to those sources’ own innate properties. Finally, digital databases also tend to be overwhelmingly large, and thus even less amenable to manual browsing than an analog archive. The ability to conduct searches through metadata thus becomes crucial to digital researchers. “Advanced Search” features on digital archives are often targeted toward specific metadata fields, allowing researchers to search (for example) for particular subjects, date ranges, and locations featured in particular records. Comprehensive application of metadata to archival materials, and consistent application of terminology within that metadata (e.g. Library of Congress Authorities for subject headings), makes such searches possible. (Vogelsang, 2024; Greer and Neatrour, 2015)

Of course, digital humanists should also bear the inherent limitations of metadata in mind. Metadata is input by individuals with their own limitations and biases, particularly when processing the records of socioeconomic groups to which they do not belong; it at all possible, it is ideal to involve donors or affected groups in the process of creating the metadata for their records, so that accurate, appropriate and respectful terminology and information is used. (Carbajal and Caswell, 2021) And since the process of applying metadata to a record is inherently subjective, variations in how one archivist versus another will describe certain records will inevitably occur, though appropriate training and best practices can at least keep these variations within a still-useable range. (Greer and Neatrour, 2015)

Tools like Tropy and Omeka can help digital humanities practitioners work with metadata. Tropy is used for initially applying metadata to primary records, including analog records that a practitioner may be digitizing themselves (e.g. photographs of documents taken in-person at a physical archive). Tropy comes with standard metadata fields already built-in to its software (e.g. Dublin Core), though these can be customized to the needs of individual resources. It also allows for the bulk application of similar metadata to a group of records, greatly speeding up the sometimes time-consuming process of applying metadata to individual records. The web-frontend software Omeka also allows for users to write metadata into the records that users upload, but one of its strongest features is its interoperability with Tropy. Records encoded with metadata on Tropy can be directly uploaded to Omeka with the metadata intact, and these in turn can be displayed publicly on Omeka with whatever fields the practitioner chooses made public.

Bibliography

Carbajal, Itza, and Michelle Caswell. “Critical Digital Archives: A Review from Archival Studies.” American Historical Review 126, 3 (2021): 1102-1120.

Martin, Greer and Anna Neatrour. “Using Metadata to Describe Digital Content.” Public Library Partnership Program, Digital Public Library of America, 2015.

Vogelsang, Henning von. “Categories Or Tagging? Differences In Taxonomy And IA.” Bootcamp, November 22, 2024. https://medium.com/design-bootcamp/categories-or-tagging-differences-in-taxonomy-and-ia-b9944ee73da8

Creating data from humanities sources – if done with care – can enable scholars to identify and potentially answer a number of questions that purely qualitative analyses alone cannot (or at least, for which qualitative analyses are less well-suited). With an extensive enough data set, humanities researchers can engage in statistical, networking, and mapping analyses that can identify and evidence broad patterns not readily apparent in the smaller data collections with which qualitative studies are typically engaged. These patterns will still need to be placed in the context of the scholar’s nuanced knowledge of their subject area in order to be meaningful, but the impetus and clarity provided by digital data can be very helpful. (Posner, 2015)

For instance, the “Tudor Networks of Power” project was able to map networks of official influence across Britain and Western Europe during the Tudor era using data relating to the velocity (i.e. frequency) of royal and diplomatic correspondence. The data was derived from the Tudor State Papers collection, which had been digitally archived via the State Papers Online but not transformed into data models usable for large-scale analysis. By closely examining and disambiguating XML data related to the collection, the project scholars were able to identify at least 20,566 unique individuals corresponding across 5,409 unique geographic locations; this data was then used to map the frequency of correspondence between these nodes. By viewing these networks, researchers were then able to pose new questions, or even tentatively suggest hypotheses regarding the exercise of power and influence across the Tudor kingdom. (Ahnert and Ahnert, 2023)

Similarly, in our own class project tracking various data points across the narratives of former American slaves (or, more ambitiously, in the similar project of Paul Escott examining a much larger data set), we were able to begin searching for patterns and correlations among the life experiences of various former slaves across the American South. While smaller-scale comparative studies such are common to comparative historical research, having a large data set of multiple former slave narratives, appropriately coded for particular data models, can allow researchers to identify larger-scale patterns not readily apparent (or at least statistically demonstrable) from smaller-scale studies.

Of course, as with any argument, the conclusions will only be reliable if they are derived from valid premises. In digital humanities data analysis, care must be taken along each step in gathering sources, compiling them into data models, and drawing appropriate conclusions. The process of transforming sources into data is not merely mechanical but a fundamentally interpretive one. Even before the scholar has begun gathering their sources, crucial decisions may already have been made for them decades or even centuries earlier – choices such as what records to retain for posterity, what types of information to include in those records, how to organize those records, and even where to house them (collections can be broken up by intent or happenstance over the years). Gaps and biases in the very creation and assembling of the original sources must always be taken into account. (Drucker 2021) If working with digital sources, the scholar must also be mindful that only a small selection of records tend to be digitized, and the decision of which records to prioritize for the digitization process is inherently determined by the needs of administrators and donors, i.e. it is a political decision. And at the point of converting sources in data points, scholars are confronted with innumerable interpretive and even ethical decisions. Given the vast amount of possible data points in, say, just a single former slave narrative, which points will the scholar choose to code into their spreadsheet (or other coding medium of choice)? How will the messiness and ambiguity of reality be reduced into a relatively small range of fixed values conducive to modeling, or even a binary TRUE/FALSE field? Should the “raw” data found in the sources be taken at face value, or should the scholar exercise discretion to account for possible biases or inaccuracies at the point of the record’s creation?

These considerations by no means render the practice of digital humanities data analysis too unreliable for use, but they should warn us against undue faith in the “neutrality” of seemingly mechanical or purely mathematical models. The best practice for humanities scholars engaged in digital data analysis is to make transparent the various points along the chain where distortions may have crept in. What process was used to create the “raw” records in the first place – whose viewpoint do they reflect, what power relationships were involved in their creation, and were they subject to alteration before being archived in their “final” form? What records were not created and therefore did not find their way into this collection? How complete was the original archive, and what gaps in the records exist? (Hoekstra and Koolen, 2019). What choices did the project scholars make when selecting their raw data, and what was the reasoning for the data models they created from it? What was their approach to disambiguating the inevitable points of confusion in any large data set? How should the arguments we derive from this data be limited or qualified by our answers to these other questions? Any responsible use of data analysis in the digital humanities will bear all of these considerations in mind.

Bibliography

Ahnert, Ruth and Sebastian Ahnert “Tudor Letters in the Digital Age.” In Tudor Networks of Power. Oxford: Oxford University Press, 2023. https://academic-oup-com.mutex.gmu.edu/book/51646/chapter/419632240.

Drucker, Johanna. “Data Modeling and Use.” In The Digital Humanities Coursebook: An Introduction to Digital Methods for Research and Scholarship, 1st ed. Taylor and Francis, 2021.

Hoekstra, Rick, and Marijn Koolen. “Data scopes for digital history research.” Historical Methods: A Journal of Quantitative and Interdisciplinary History 52, 2 (2019): 79-94.

Posner, Miriam. “Humanities Data: A Necessary Contradiction.” Miriam Posner’s Blog, June 25, 2015. https://miriamposner.com/blog/humanities-data-a-necessary-contradiction/.

Though digitization can play a very important role in helping to preserve records and disseminate them more widely with various audiences (Terras 2012,, 1-3; Terras 2022, 257), it also has real limitations of which to be aware. Beyond the macro-level consideration that only a small fraction of physical records will realistically be digitized in the near future (given financial, logistical, and political considerations of holding institutions and their funders; Terras 2012, 3-4; Terras 2022, 257), there are also limitations inherent in the digitization process itself.

To put it simply, a digital facsimile will never be a perfect copy of any physical item. The degree to which it diverges, however, will depend greatly upon the nature of the original item or record. For written records like documents, digitization can actually capture a great deal of the content of that record. (That is, assuming the digitization is done with care – poor lighting, bad focus, or lossy compression can greatly reduce the legibility of a scanned document). The same is largely true of digital copies of flat artworks like posters, flyers, or even paintings, assuming that the hardware used to make the digitization is of sufficiently high-quality to capture the nuances of coloring and shading. Even here, however, a high-quality digitization will never be able to fully or even adequately capture the tactile qualities of a basic document, much less an artwork that may have subtle bumps or scratches. Nor can it give a real sense of the relative size of a document or piece of art; one looking at a digital copy of Picasso’s Guernica on a standard computer screen, for example, would hardly know that the original painting takes up nearly an entire wall in a museum. And of course, digital scans can only capture one side of a flat object at a time, though housing scans of each side together in the same location can mitigate this limitation.

The divergence from the original source material becomes even more dramatic as we move from flat items to those in three dimensions. Two-dimensional photographs of three-dimensional objects, no matter how high their quality, may do an adequate job of some of an object’s visual quality never be able to give a complete sense of an object’s shape or how it feels to the touch. Three-dimensional scans can render physical objects as manipulable digital objects, allowing users to view all sides of an object as though they were examining it in-person, though the quality of the scan depends greatly upon the sophistication of the equipment (including software) as well as the individual skill of the scanner. But even the most skilled 3-D scanning technician will be unable to render any sounds that an object might make, much less recreate the interactive element of objects that are meant to be used rather than just seen.

These limits of digital imaging became evident during the kitchen exercise that we performed in this module. Even digitizing my family’s simple shopping list proved less straightforward than I had initially assumed, once I realized from doing the exercise that the feeling, weight, size, and back side of the list would impossible to capture. Perhaps feeling and weight are not such crucial elements when one is dealing simply with documents, but the limits of digitization were brought home quickly when I moved on to three-dimensional objects.

Working with digitized representations can have a significant impact how we understand different kinds of items. Digitization can have a flattening effect in the sense of increasing the sense of likeness between resources that have been stripped of many of their physical qualities such as size and function. Given the relatively small number of resources that have been digitized, an outsized focus on digitized items may artificially inflate their importance relative to non-digitized resources and/or obscure the significance of those items (and the narratives they convey) that have not been digitized. Digital objects tend to be de-contextualized, in the sense that they are discoverable via searches that remove them from their archival context and thus render their relationship to other records more obscure.

At the same time, digitization also has many advantage for our ability to use resources for different purposes. The availability of digital records online allows researchers and the general public much easier access to records they would otherwise have to travel long distances to consult. Digitalization also facilitates the creation of large data sets that can be used for various research initiatives such as statistical modeling and text mining. 3-D visualizations allow for access to artifacts or sites that may be restricted or inaccessible due to preservation concerns, as well as for the virtual reconstruction of damaged or degraded artifacts (Drucker 2021, 163-166).

Bibliography

Drucker, Joanna. “Photogrammetry.” In The Digital Humanities Coursebook: An Introduction to Digital Methods for Research and Scholarship, 1st ed. Taylor and Francis, 2021, 162-168.

Terras, Melissa. “Digitisation and Digital Resources in the Humanities.” In Digital Humanities in Practice, edited by Claire Warwick, Melissa Terras, and Julianne Nyhan. Facet Publishing, 2012.

—. “Digital humanities and Digitised Cultural Heritage.” In James O’Sullivan (ed.), The Bloomsbury Handbook to the Digital Humanities. London: Bloomsbury Academic, 2022, 255-266.

Locating digitized primary sources for a research project can take a bit of finesse. Digitized primary sources tend to be located in collections online, which can be found in various ways.

If you have access to a research library, it will subscribe to many different kinds of databases that contain collections of primary sources. Some of these databases will be quite broad, e.g. collections of major national and regional newspapers. Others can be quite narrow and topic-specific. The George Mason University Library, for example, has primary source collections dealing with historical topics ranging from The Cold War to the African American Civil Rights Movement.

Other useful databases can be located online and may or may not require a subscription or institutional affiliation. Large public entities like the New York Public Library and the Smithsonian Institution maintain online databases of primary sources. Again, each of these databases may contain particular collections of relevance to your research subject; navigating the website and looking for a tab called “Collections” is often a good place to start.

Once you have located a relevant online collection, you are ready to begin searching. Some online collections are sorted into pre-existing categories, often arranged by theme, time period, or the name of notable individuals. You can use these pre-set categories to browse the collection.

However, most collections are too large to examine all of their records individually, even when browsing by category. In addition, one of the major advantages of using a digital collection is that they usually include a search function. The search function not only makes it easier to locate records relevant to your project, but they can also be used to bypass the pre-set browsing categories that may or may not be in accordance with your interests or topic. This is especially true if you are researching a topic that is off the beaten research path, or if your subject may involve coded terms with which professional archivists may not be familiar. (Frydman 2025).

So it is helpful to come prepared with a series of keywords or phrases to search. Be aware that not all search bars are created equal, particularly if you are searching for multi-word phrases rather than individual words. It is helpful to use the “Advanced Search” feature if one is available on the particular search bar that you are using. For example, conducting a basic search for the phrase “George Washington” (in quotation marks) may or may not result in records that contain only the word “George” or “Washington,” depending on how the collection you are using configures its basic search. Using the Advanced Search, however, one can conduct an “exact phrase” search for only records containing the words “George Washington” in sequence, giving more relevant search results.

By way of illustration, I was recently conducting an online search for primary sources relating to the Chicano movement in Sacramento, California. While some collections online specialize in Chicano materials, the vast majority do not. When using the collections that weren’t specific to Chicano history, I began my search – naturally enough – with the term “chicano.” My aim here was to start out with the widest possible universe of potentially relevant records, as I know from my prior experience with the subject that pretty much any primary source article or interview dealing with Chicano politics would likely include that word somewhere, either in the publication name or the article itself. The initial search results might still be a bit too large to comb through one-by-one, but it would give me a narrower base upon which to use more targeted search terms.

Using the general “Independent Voices” periodicals collection on JSTOR, for example, gave me 4,435 results for the search term “chicano.” For further refinement, one helpful starting point was to use filters to narrow the results based upon the time frame that I was most interested in. For my case, I knew that any records earlier than 1960 would likely be irrelevant to my search, since my subject matter did not become a political term of reference until the 1960s. By contrast, when searching the Sacramento Movimiento Chicano and Mexican American Education Oral Histories collection, which contains transcripts of interviews with Chicano activists, I did not start by searching for “chicano” because I knew that term would likely already be included in the vast majority of records in the collection.

Starting from this narrower base, it then helps to search for more specific terms, based upon your research topic. Using Boolean connectors is helpful. For example, if I am looking for records relating to the murals in Chicano Park in San Diego, I can specify in the search that I am searching for records containing the term “mural” AND the term “park” AND the term “San Diego.” Prior research into your subject can inform the particular search terms you will use, particularly if you are researching communities that use a vernacular not widely known or used by the professional archival community.

So I found a really cool photo online that I want to use in my own project. Can I just go ahead and do that?

WAIT! Just because you found something while looking around online doesn’t necessarily mean you can use it any way you want, if at all. Have you checked its copyright status?

Have I checked the what?

Copyright status. It refers to the legal power that the creator of something has to determine who – if anyone – is allowed to use, share, or reproduce the thing that they created. (Technically, it’s a property right. And like most property rights, it can be transferred or loaned to other people who then become the new copyright holders or licensees.)

I thought that just applied to using Disney characters and other famous stuff.

Nope! Pretty much anything that someone has put their own original thought or talent into might be subject to copyright. It doesn’t even need to have been published.

What do you mean “published”? It’s just someone’s family photo from a few decades ago that I found on an archive online. Why would that be copyrighted?

I’m not saying it’s definitely copyrighted, I’m just saying you need to check first. And “published” doesn’t necessarily mean published in a book or a magazine, it just refers to when something was distributed publicly, like putting up a poster or performing a play. But even something that was meant to be totally private can still be copyrighted, as long as someone put their own thought or talent into making it.

I’m pretty sure this is private, it’s just an old photo of someone’s family from like a hundred years ago.

Do you know who the photographer was? If so, depending how old the photo is, the copyright may have expired if they’ve been dead for 70 years.

No, I don’t know who the photographer is…

Well, is the photo very old? Because even if you don’t know the photographer, the copyright may have expired if its an unpublished work and it’s been 120 years since the photo was taken.

I have no idea how old it is, but I don’t think it’s that old –

OK, is there a section on the web page where you found the photo that talks about “copyright”?

I don’t know, I just see a photo. Where would I find that?

Well, every web page is different, so there isn’t necessarily one place you would definitely find it. But try scrolling down (or doing a CTRL+F search for “copyright”).

OK I’m scrolling down. Yeah there’s a bunch of fine print here that I didn’t really care about. Let’s see…

Look for a field called something like “Copyright” or “Usage.”

Oh wait, I found it! “This item protected by copyright and/or related rights.” Oh no! Does that mean I can’t use it after all?

Not necessarily. You could look up the copyright holder and ask for permission to use it, though that could take a long time and you don’t know what they’ll say. Plus, you may have to pay money for it.

Yeah, I don’t want to wait that long, plus I don’t have any money.

Well, there is another possibility. Depending on the type of project you’re doing and how you want to use this photo, you might still be able to use it under the doctrine of “fair use.”

Sweet, now we’re talking! But uh, what does that mean?

“Fair use” is a legal term describing circumstances under which a copyrighted image or other record can still be used without violating copyright.

That sounds complicated. How do I know if what I’m doing is fair use?

It’s not actually as complicated as it sounds. But unfortunately there’s isn’t a bright-line rule, either. The basic consideration is whether your use of the photo would realistically impact the market for other people to pay the photographer for the use of the photo. So, you may not be able to use the entire photo in your project, since other people could then just screengrab it from your project, but maybe you could just use a smaller part of the photo? Or maybe you’re planning on altering the photo in some fundamental way? (This is what’s called a “transformational” use). It also depends whether your project is commercial or not, as opposed to, say, educational. Not that you can’t ever have fair use in a commercial project, but it weighs against you. These are all factors to balance, rather than an absolute checklist.

I’d rather not deal with all that, plus I wanted to use the whole photo in my commercial documentary. I’d rather just look for another one. Let’s see… OK, here’s one that might work. I’ll scroll down and see what the copyright status is… Hmm, it says “this item is in the public domain.”

That’s great! It means you can use the photo in any way, without restriction.

Sweet! Let’s find another one just for good measure… Here’s another good one… Oh man, it says “made available for non-commercial, informational, and educational purposes under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.” What the heck does that mean?

It means that it’s under copyright, but the owner is allowing people to use it in their projects anyway, so long as the project is educational rather than commercial, the photo isn’t altered, and the photographer is given credit. People use Creative Commons or other kinds of licenses to let people use copyrighted images or records under certain terms of use.

Terms of what?

“Terms of use.” A lot of the images or other things you want to use might have a section underneath them titled “Usage” or “Terms of Use” that lists whether a copyrighted image is available for certain types of use. If you don’t see one, you can always go to the main page of whatever website you’re using and see if the entire website has a Terms of Use section.

I see a lot of these images have “CC” information, but I don’t recognize all the acronyms.

“CC” stands for Creative Commons. The other acronyms stand for the different kinds of licenses that the copyright holder is giving you. It’s a lot to memorize; you can use this website to quickly look them up.

This is all a lot to keep in mind.

Yes, that’s true. If you want to play it safe, a lot of sites where you search for images or other records allow you to filter the results by copyright status. Or you can use a database that you know is filled with public domain or Creative Commons-licensed materials, like the Public Domain Image Archive, Getty Open Content Program, the Prelinger Archive.

A Google search for “public domain images” or “Creative commons images” should give you a list of such sites — but still remember to double-check the copyright and terms of use on the individual image page or, failing that, the site-wide terms of use page!