This study treats finding aids as historical objects of rich textual data that will be analyzed through “tagging,” a method of qualitative data analysis often utilized in social science fields. Also known as “coding” or “labeling,” tagging allows the researcher to identify concepts, themes, opinions, and sentiments across a collection of narrative data which may then be analyzed to determine patterns such as frequency and co-occurrence of tags across the corpus of data.

I gathered a corpus of 78 finding aids from TARO, a database of thousands of EAD finding aids contributed by over 80 archives and libraries across the state since 2001.¹ I compiled this collection of finding aids by searching the following keywords that I expected might be included in records related to the racial caste system in Spanish colonial society: mestizo, mestiza, mulatto, mulatta, castizo, castiza, negro, negra, indio, india, Indian, creole, criollo, criolla, caste, and casta. I also searched categorical terms that I expected might have been added to these records’ descriptive metadata as subject headings or mentioned in historical/biographical and scope and content notes required by DACS (Describing Archives: A Content Standard): mixed race, race relations, interracial, biracial, and miscegenation. I limited the results of these keyword searches to those with start dates before 1900, since this study is focusing on the keywords’ semantic meanings during and immediately after the colonial period and excludes analysis of mestizaje as it was framed in the 20th century. Although racial caste terms were no longer employed by Mexico after it achieved its independence in 1821, I also wanted to account for any potential remnants of colonial perspectives on race that may have endured into the 19th century and became intertwined with Anglo-American conceptions of race as the latter sought to supplant Mexican hegemony during their colonization projects in Texas the Southwest US.

After gathering over 100 finding aids based on these keyword searches, I navigated to the locations in the text where the keywords were found and removed finding aids in which the keyword was within the title of a written or creative work, a person’s name, or the name of a geographic location. Once the corpus was culled, I uploaded the finding aids as PDF documents to ATLAS.ti, a qualitative data analysis software typically used for coding and analyzing large bodies of textual documents such as transcripts & field notes. I developed two series of codes: the first series of codes are intended to flag racial descriptive terms that appear anywhere in the finding aid (Table 1). In addition to coding descriptions of mixed-race subjects, I also coded terms used to describe Black, Indigenous, and Spanish subjects who were not identified as mixed-race in the hopes of being able to discern differences in descriptive approaches to each racial group.

Table 1: Coding description of race and caste
Narrative description	Codes
Colonial-era terms used to classify mixed-race people	mestizo(s), mestiza(s), mulatto(s), mulato(s), mulatta(s), mulata(s), castizo, castiza
Colonial-era terms used to describe racial caste as a social hierarchy or racial imaginary	caste, casta, limpieza de sangre
Common terms used to generally describe Black people in the Americas. These terms remain in use well beyond the colonial era	Black, African, negro (Spanish and English use), negra (Spanish use)
Common terms used to generally describe Indigenous peoples in the Americas. These terms remain in use well beyond the colonial era	indio(s), india, indian(s)
Colonial-era terms used to classify Spaniards born in the New world	creole, criollo, criolla
Modern phrases referring to mixed-race groups	mixed race, race relations, interracial, biracial

The second series of codes were intended to identify characteristics of the finding aids and collections they represented (Table 2). These codes were intended to help build an aggregate profile of the types of repositories, collections, and descriptive labor represented in the study. The comparison between these characteristics – especially the visibility of artificial arrangement and reparative description efforts – and the aforementioned codes for racial terminology were essential for understanding relationships between archival labor and their approaches to racial description.

Table 2: Coding general characteristics of finding aids
Metric	Codes
Evidence of archivists’ attempts to employ reparative or culturally sensitive language in order to make users aware of outdated or racist language within the collection	reparative/aware
Date description was last revised	Codes developed based on dates of most recent processing or revision listed in the finding aids. If no date was present, finding aids were coded with “date unknown”
Era in which collection inclusive dates fall	pre-independence, bulk pre-independence, revolutionary era, post-independence, bulk post-independence, both eras, no dates provided
Repository type	government, museum, public library, religious, university
Description level	series, box, folder, item, volume

---

Notes