Introduction
This article emerged from a talk I gave as part of ARLIS series on cataloguing ethics in relation to my PhD research on how North American Indigenous languages have been collected and catalogued at the British Library (BL).Footnote 1 The problems that dominant library cataloguing systems pose to Indigenous materials is well-documented: critiques are generally levelled against the terminology of the catalogue, the use of irrelevant subject lenses and the overall misrepresentation of Indigenous knowledges.Footnote 2 However, an additional layer of ethical and methodological consideration is added when those library collections exist as a geographic remove from Indigenous peoples. The BL holds a significant collection of printed materials in, and about, North American Indigenous languages that largely speaks to a history of colonial and settler-colonial projects and collecting. This article explores what that collecting context means for how we find, experience and encounter Indigenous language texts in the library today.
I begin by outlining how language materials were found at the BL, before offering an approach to the catalogue. Here, I argue that seeing the catalogue as an archive, and interrogating the relationships between texts and cataloguing systems, reveals assumptions in knowledge organisation. A contextual reading of catalogues shows institutional collecting practices; unsettles standards, authorities and terminology; and surfaces examples of absence, resistance and non-conformity.
I then apply this approach to the history of Indigenous language collecting and cataloguing by tracing a brief history of the LCSH ‘Indians of North America—Languages’, a term that reoccurs in the BL's catalogue. The harm of the continued use of this term is explained by Sarah Dupont, head librarian of the Xwi7xwa Library in British Columbia:
When Aboriginal Peoples go looking to find representations of our many diverse knowledges in the Library collections, we should be able to search using terms we use to describe ourselves. We should not feel the sting of antiquated, colonial, and racist words that perpetuate negative connotations of us, especially when we and our Allies go to do the work of lifting academia and broader society out of these shadows through our scholarly efforts.Footnote 3
Those scholarly efforts have seen both challenges to dominant systems, and the centring of Indigenous Knowledge Organisation in library practice.Footnote 4 For Indigenous languages, projects have addressed the lack of specificity in language description, as well as the impact of unintuitive linguistic and geo-spatial groupings in classifications.Footnote 5
However, I take a historic approach that specifies this to languages. Whilst the Library of Congress (LoC) began printing subject headings in the late 19th century, I follow a longer history that shows how ideas about languages have become embedded in library systems in ways that are apparent today. This history begins with Thomas Jefferson, whose ideas and methods in language collecting were influential and whose personal library formed the first classification of the LoC. It ends with John Wesley Powell's 1891 Indian Linguistic Families of America, North of Mexico as a text that epitomises extractive collecting and as the text used for the warrant in initially formalising LCSHs for languages. Powell was the first director of the Bureau of Ethnology and headed the institutionalising of salvage anthropology, the term given to early American anthropology that upheld the colonial logic that ‘vanishing’ Indigenous cultures needed preserving whilst simultaneously using the output of that assertion to assimilate and eradicate Indigenous cultures and languages. This history shows that parts of the catalogue are an artefact of problematic bodies of literature, whilst also showing that examples of Indigenous resistance can be used to reframe the catalogue.
Reading the catalogue
Searching the BL's catalogue for Indigenous language materials today intersects with various cataloguing contexts. Items can be located using LCSHs and MARC language codes, which are three-letter alphabetic strings from a controlled list that identify languages. These codes cover languages - for example [moh] for ‘Mohawk’- language groups, [iro] for ‘Iroquoian (Other)’, and a broad, collective code - [nai] ‘North American Indian (Other)’.Footnote 6 The BL, which was the British Museum Library (BML) until 1973, adopted the use of LCSHs in the late 20th century. The BML followed an internal classification and cataloguing process, and this history means many cataloguing records are incomplete today. Between 1880 and 1881, the BML began printing an author-title catalogue and a subject-index, and subsequent catalogues are an entry point to the BL collections, and an understanding of historic subject arrangement at the BML. Therefore, retrieving records with a relevant LCSH or MARC code, running author-title searches, comparing with other catalogues and referring to the BML's catalogues generated a dataset of materials in, or partially in, North American Indigenous languages, published after 1850. This dataset holds over 500 materials representing over 50 languages and is weighted towards missionary, ethnographic, anthropologic and non-Indigenous authored linguistic material.
Viewing catalogues as archives to be read, and places to enact change, forms an approach to the BL dataset. A similar view is taken by Melissa Adler who ‘reads the library shelves and catalog as primary sources’ to trace a history of sexuality and categories of perversion through the LoC's cataloguing practices.Footnote 7 Adler's work shows a relationship between bodies of literature and cataloguing standards, where LCC and LCSH both ‘give form to the academic disciplines’ and ‘actively produce, reproduce, and privilege certain subjects and disciplinary norms.’Footnote 8 This reciprocation between disciplinary knowledge and organisational systems enforces categories, embeds assumptions and makes claims as to what does, and does not, constitute knowledge. Therefore, to read the catalogue is also to read a narrative of loss for the ‘lives of people who have been silenced and denied access in the past, those whose lives were abstracted, catalogued, and classified; and for the continued marginalization of subjects’.Footnote 9 Adler's insights can be extended, where today, of 574 federally recognised tribes, the majority have experienced, or are battling, language loss. A collection composed of anthropological projects that objectified languages; and missionary, boarding school and U.S government presses that aimed to ‘civilise’ and eradicate languages certainly reads as a catalogue of violence and absence.
Exploring the relationship between texts, and the catalogue, can offer insight into the construction and limitations of the catalogue. Some relationships are in tension. In a recent critique by Lee, Bullard and Dupont they use what they term ‘meta-texts’ which are ‘titles about the particular problems surrounding the representation of Indigenous knowledge and communities in the English language.’Footnote 10 For example, the use of ‘Indians of North America—Canada’ to describe Greg Younging's Elements of Indigenous Style: A Guide for Writing by and about Indigenous People, a text that argues ‘Indian’ is outdated and problematic, is particularly jarring. Yet, some texts appear to resonate with the catalogue. In the following section, I trace the connections between Powell's Indian Linguistic Families of America, North of Mexico, the work of the Bureau of Ethnology and the development of subject headings at the LoC. Here, it is the decided lack of tensions between those works and the catalogue that raises ethical concerns.
Some texts reveal that, despite the reality of loss, catalogues can also be read for resistance and non-conformity in ways that reframe absence. When starting to write Common Pot, a book that explores Native writing traditions in the Northeast, Abenaki scholar Lisa Brooks writes:
When I first began, I was told that looking for writing by Indians would be like looking for needles in a haystack. But I figured if you know the names of the needles and the places they are from, it might be easier to find them.Footnote 11
Indigenous presence can always be found if we step outside of prevailing assumptions and systems. The text is prefaced with a definition of native writing as both a tool and a process: writing is active, place-based and participatory. This conception of writing is mirrored in the nature of the Abenaki language, a language that, like many other Indigenous languages, is composed of verbs, movement and relationality. Brooks defines her book as ‘at once an activity in which we participate, an instrument, and a map’, a map that asks us to consider the print output of Anglo-American traditions as instead ‘participants in Native space’.Footnote 12 The book also critiques ideas of authenticity in defining what does, and does not, constitute native writing, where Indigenous authored materials in newer languages, like English, also transmit linguistic and cultural knowledge. Here, we might ask what kind of catalogue would emerge from this conception of Indigenous writing and languages? What would a catalogue that presents Indigenous writing as adaptation and independence, that holds networks and that expands a definition of ‘language material’?
Indians of North America—Languages
Thomas Jefferson's connection to the Library of Congress (LoC) and his wider ideas about Indigenous languages offer a starting point for the intersection of languages and cataloguing. In 1814, the LoC acquired Jefferson's personal library of 6,700 books as well as his classified catalogue. Composed of 44 ‘chapters’, or subjects, this classification formed the initial organisation of the LoC. Indigenous languages were an afterthought in Jefferson's catalogue: he owned only two books- Zeisberger's Delaware-Indian Spelling Book and a Mohawk Liturgy- which appeared in chapter 43, ‘Criticism. Languages’.Footnote 13 This does not reflect Jefferson's personal interest in languages as, between 1790 and 1810, he collected over 400 vocabularies for his ‘Indian Vocabulary Project’ in an attempt to classify all Indigenous languages. He believed that vanishing languages needed to be quickly collected, and he also saw languages as a source to reconstruct the history of Indigenous groups in America, and of human origins. On this view, languages became sources that served a narrative of American antiquity, the scientific advancement of a modern America and a flawed aspiration to universal knowledge.
The materiality of Jefferson's project is also significant in demonstrating the use of lists, tables and catalogues to objectify and order Indigenous languages. In collecting languages, Jefferson circulated broadsheets of standardised word lists following a uniform orthography to collectors. As with the methods of Powell's later work, the use of lists was not harmless: they meant the literal extraction, reordering and recoding of languages in archival and library settings, often for the purposes of forcibly removing peoples from their ancestral lands. Yet, Jefferson's project did not work as he intended. Jefferson's vocabularies today offer instead examples of resistance and refusal and are used by communities today to reclaim and preserve their traditions.Footnote 14
By 1897 Jefferson's classification was unworkable, and the LoC began mapping a new classification from his original. In 1902 the LoC also began printing and exchanging subject headings with other libraries, a process that accelerated into the 20th century and constituted the origin of copy cataloguing. In the first and second edition of Subject headings used in the dictionary catalogues of the Library of Congress the entry ‘Indians of North America—Languages’ referred readers to John Wesley Powell's Indian Linguistic Families of America, North of Mexico printed in 1891 as the standard authority on ‘the names of Indian linguistic families.’ Under this heading, we find the classification and names of Powell's work. Even under the general entry ‘Indians of North America’, Powell's linguistic families are used to ethnologically classify tribes, an entry that under ‘culture’ includes the redirection ‘for books treating of efforts to civilize the race see Indians of North America, Civilization of’.Footnote 15 Already, the narrative of America in the subject headings authors a narrative of Indigenous disappearance reflected in the systematic grouping of peoples by language.
To understand the implications of Powell's work as an authority for the LoC, we must explore the rationale, method and context of his classification. Firstly, Powell's work emerged from Powell's position as the first director of the Bureau of Ethnology, a branch of the Smithsonian Institution established by an Act of Congress in 1879 committed to the concerted collecting of North American Indigenous cultures. Languages were at the centre of the Bureau's ethnological and anthropologic project. In the First Annual Report of the Bureau, Powell outlined his systematic plan to classify all languages, to map the ‘habitats’ of linguistic groups and to establish consensus on terminology for languages. Languages were the ‘anthropological key’, where ‘customs, laws, government, institutions, mythologies, religions, and even arts cannot be properly understood without a fundamental knowledge of the languages which express the ideas and thoughts embodied therein’.Footnote 16 Whilst a longer history of classification is relayed here, the late 19th to early 20th century saw a vast acceleration in the collecting of, and institutionalising of, Indigenous cultures, the catastrophic legacies of which still impacts how collections are found and used today.
In 1877 and 1880, Powell distributed his Introduction to the Study of Indian Languages to gather the information for his classification. The Introduction directed active and amateur collectors on the proper recording of languages in the field to ensure that the data was usable for comparative study. The workbook was divided into 29 ‘schedules’ designed to compartmentalise Indigenous life with each section consisting of ‘words, sentences and phrases’ for collection. From ‘persons’, ‘parts of the body’ and ‘implements’ to ‘divisions of time’ and an extensive schedule on relationships and kinship relations, the schedules indicated yet another method of extracting and restricting Indigenous words through predetermined and often irrelevant topics. Hundreds of these workbooks were completed across the North America by Bureau employees, other ethnologists, linguists, missionaries and army personnel and sent back to the Bureau in Washington D.C. A blank box at the beginning of each workbook which required the collector to state ‘tribe’, ‘locality’, ‘recorded by’ and ‘date of record’ informed the cataloguing of the manuscripts in the archive.Footnote 17 These workbooks were then used to order languages in the creation of the accompanying linguistic map.
From the anthropological drive to classify, to the use of standardised and reductive methods that erased Indigenous perspectives, the process of creating the classification encapsulates many issues still encountered in catalogues today. The collector is prioritised in the workbooks, and in the cataloguing of the workbooks in the archive, yet the completed books relied on Indigenous input. Many who completed the workbooks produced other publications on Indigenous languages, with their work often printed in Bureau reports, demonstrating a centralised print network. The completed workbooks hold examples of interaction, of the two-way process of translation, and of the frustration of the collectors at restrictive format. In many of the books, Indigenous names, place-names and questions litter the margins, and in others the pages and pages of unfilled boxes speak to another type of interaction, or lack of. These realities are flattened in the classification and then used by the LoC to describe, order and define books. This shows both the reciprocation of knowledge and terms between the Bureau and libraries, but it also historically locates this process. In doing so, it opens us up to alternative histories, and alternative ways of organising.
The history presented here is incomplete. LCSHs have undergone various updates and there are other approaches and bodies of knowledge to consider. However, catalogues like the BL's continue to use ‘Indians of North America—Languages’ and non-Indigenous naming practices for languages. Whilst the BL adopted LCSHs in the late 20th century, the BML's history still speaks to this history. The print output of the Bureau clearly informed subject indexing at BML in the late 19th century. In 1886, Indigenous languages are located under the subject entry ‘INDIANS, American’ subheading ‘Languages’ under which there are four entries. The first of these entries is a blank copy of Powell's Introduction which was donated to the BML on 28th February 1880 by the United States Department of the Interior.Footnote 18 Even in the 1970's the subject entry remains consistent: ‘For the American Indians, see INDIANS’, ‘Languages of the Indians of Canada and the United States’. Readers are also directed to ‘INDIANS’ for the ‘Antiquities’ of America, and Indigenous Art redirects to ‘Ancient, Art of Prehistoric and Primitive Races’. Further language material can be found under ‘Ethnology’, and ‘Anthropology’ reflecting the view of Indigenous languages as source material for disciplinary knowledge.Footnote 19
Conclusion
A study of the LCSH ‘Indians of North American—Languages’ shows the ways histories of language collecting emerge in how materials are found today. From Jefferson to Powell, this LCSH speaks to a history of ideas and methods that attempted to objectify and erase languages. Whilst the term is problematic, it also relates to wider cataloguing issues. The generalising or miscataloguing of language materials, the format and materiality of the catalogue and the formalising of a non-Indigenous definition of language materials speaks to the same treatment of languages. It returns us to the harm described in the introduction that occurs when peoples cannot access materials in ways that represent themselves. Yet, this history is also a history of Indigenous resistance, and these examples can be brought as a challenge to the catalogue.
The history offered here is partial: there are other bodies of literature and texts that could be used to interrogate ideas around languages and knowledge organisation in libraries. However, there is a wider methodological claim here to see catalogues as active in how collections have been built and made knowable. By surfacing the tangled and contested nature of language classification, and different conceptions of writing, language and organisation, we can envisage other futures. Holding texts and the catalogue in a relationship makes clear assumptions in knowledge organisation that raises further questions. Can the catalogue recontextualise materials, or recognise its own history as an artefact of knowledge organisation? Can, or should, catalogues subvert the original intentions of materials to hold these histories of resistance? Perhaps that is not feasible, or desirable, for the catalogue. However, such questions are important as libraries consider their responsibilities in representing North American Indigenous languages.
