Please enable JavaScript in your browser.

Bryan Zehngut-Willits

Metadata Methods: Working Against Bureaucratic Categories of Exclusion

Published on

The Keeping Records and the Golden Gate team acknowledges that archives and collections, including our own, are sites of power that have the potential to assign categories in their metadata schema and ontology that are not only inaccurate, but potentially harmful. We designed this project to reveal these categories as much as possible without replicating them. To this end, we employed several methods when designing our metadata schema that work against the government’s intention of assigning migrants to specific categories for the sake of restriction and exclusion.

First, our metadata is derived from each individual document so that categories like race, color or sex, for example, are not being replicated in places other than where they already exist in the file. Take for example Alexander Ostroumoff’s A-File, which contains Form N-300 from the INS. This form requires Ostroumoff to declare his sex, color, complexion, race, and nationality, among other things. Because these categorizations exist on this particular document, we reflect that in the metadata.

If we were to build our schema around the entire A-File, however, that would mean attaching the government’s racial and color designations to all the documents in Ostroumoff’s file, thereby applying these categories to documents that otherwise do not contain them. This, in effect, would reinforce the power of these invented categories. On forms like the I-144, therefore, we do not record metadata categories like race, color, etc. since they do not appear on the document itself.

Another way to work around replicating categorizations while still identifying them in your metadata would be to use an explicit metadata field like “government assigned race,” instead of having a field called “race.” This might work as a kind of disclaimer to express an understanding that these categorizations are invented ones that immigration officials use for their own ends. Many fields in our metadata schema could ostensibly begin with “government assigned,” but the need for this redundancy is alleviated by working on the document level, rather than on the level of the A-File and the individual to whom it pertains. In other words, our method expresses that “the government assigned Ostroumoff to category X on this document,” rather than “Ostroumoff is category X.”

In some cases, too, it is every bit as telling to note what data is not being collected. Why the government might want to know an immigrant’s economic status or country of origin on one form at one point in time, but not at another, may have a significance that would otherwise be obscured if all the documents in an A-FIle were given the same metadata values simply because those values appeared at least once in the file. Recording categories in our metadata only for documents where they appear has the additional benefit of keeping the fields of our dataset cleaner and more manageable.

Next, when it comes to government forms that must be filled out and signed by non-citizens, we designate the government official who witnessed or cosigned the form as the principal creator of the document. In most cases, the immigration official is filling out forms with data the migrant provided in writing or through oral testimony, as sometimes took place in an interview at a border crossing checkpoint or internment camp. Accordingly, the creator of Tora Matsouka’s Application for a Reentry Permit, Form 631, is the immigrant inspector, William Thomas, who signed the form and presumably filled it out via typewriter.

In some cases, migrants or non-citizens applying for a visa, naturalization, or permission to travel across US borders provided the majority of the information and performed most of the actual filling out of a form. We can see that, for example, with Naoko Takita’s Application for Certificate of Identification (Aliens of Enemy Nationalities), form AR-AE-22

Even though Takita provided the information and presumably filled out the form, she did so at the behest of the government. So if the signature or mark of some immigration officer is present on a form that a migrant had clearly filled out themselves, we still consider the immigration officer signing off on the form as the primary creator. The creator in this case, then, is the identification official, Bruce Moffitt. By designing our metadata schema in these ways, we further underscore how the bureaucratic practices of immigration and border control are responsible for extracting information from migrants that can be used to assign them to categories that may result in their exclusion, expulsion, observation, or incarceration.