African Literary Metadata and Makerere University's Library

Abstract

While Africanist librarians have often pointed out the inadequacies of the international library classification system for the cataloging of African materials, subject specialists in African literature have paid little attention to the problem of how African literatures are cataloged in the world's leading classification systems. This is so despite the fact that diverse modalities of African literary production (from oral forms, to ephemeral print literatures and pamphlets, to a variety of Web 2.0 forms) sit uncomfortably in existing classification schemes, just as they have troubled hermeneutic and analytical methodologies in the field of African literature for decades. This article provides a case study of the history of the cataloging system at Makerere University Library and discusses how this has come to shape the body of African literature housed there, even to this day. I focus on Makerere University because of its key position in debates and discussions about the politics of Anglophone African literature in the 1960s, as well as the fact that it was the training ground of major African literary scholars, activists, writers, and educators. This makes it the perfect microcosm through which to think about the role of cataloging systems in the structuring of disciplinary and political knowledge. The article focuses on African Literature in the Dewey Decimal Classification system and the Library of Congress Subject Catalogues and then close-reads the library in Ngügi wa Thiong'o's A Grain of Wheat by way of showing how imbricated library classification was in colonial ontologies and how this played out, not only in the placement of literature on library shelves, but also in literary content.

Headnote

ABSTRACT

While Africanist librarians have often pointed out the inadequacies of the international library classification system for the cataloging of African materials, subject specialists in African literature have paid little attention to the problem of how African literatures are cataloged in the world's leading classification systems. This is so despite the fact that diverse modalities of African literary production (from oral forms, to ephemeral print literatures and pamphlets, to a variety of Web 2.0 forms) sit uncomfortably in existing classification schemes, just as they have troubled hermeneutic and analytical methodologies in the field of African literature for decades. This article provides a case study of the history of the cataloging system at Makerere University Library and discusses how this has come to shape the body of African literature housed there, even to this day. I focus on Makerere University because of its key position in debates and discussions about the politics of Anglophone African literature in the 1960s, as well as the fact that it was the training ground of major African literary scholars, activists, writers, and educators. This makes it the perfect microcosm through which to think about the role of cataloging systems in the structuring of disciplinary and political knowledge. The article focuses on African Literature in the Dewey Decimal Classification system and the Library of Congress Subject Catalogues and then close-reads the library in Ngügi wa Thiong'o's A Grain of Wheat by way of showing how imbricated library classification was in colonial ontologies and how this played out, not only in the placement of literature on library shelves, but also in literary content.

classification systems to order and make visible and accessible African Le scientists have long lamented the inadequacy of international library materials and bodies of knowledge (Kisiedu; Kotei; Kyle; Mowery). Debates have focused on the inadequacies of international classification systems to manage African names (Mutula and Tsvakai) and languages (Bein), as well as the tenacity of race-based classifications of people in these systems (Furner). More recently, library and classification science has been substantially opened and enriched by discussions as to how indigenous forms of knowledge can be both included in, and themselves transform, knowledge ontologies (Littletree and Aden, Schweitzer and Henry, and Wisecup). Despite this rich body of work, subject specialists in African literature have paid little attention to the problem of how African literatures are cataloged in the world's leading classification systems. This is so despite the fact that diverse modalities of African literary production (from oral forms, to ephemeral print literatures and pamphlets, to a host of Web 2.0 forms) sit uncomfortably in existing classification schemes, just as they have troubled hermeneutic and analytical methodologies in the field of African literature for decades.! In this article I argue that library cataloging norms have contributed to the structural invisibility and cultural devaluation of African literature in three primary ways: first, by separating orature from literature; second, by prioritizing published book forms over other kinds of print; and, third, by enabling the heightened visibility of African writing in colonial languages and concomitantly obscuring the visibility of literature in African languages.

While 1 follow Karin Barber in my commitment to taking "oral and written traditions as a unified field of inquiry" (29), I wish to investigate rather than dismiss what she sees as the "culturally-loaded category" (29) of "literature." I do this because, as this paper seeks to outline, this very "cultural-loading" has resulted in the devaluing of the enormous body of African created literary and oratory expressions, what Barber prefers to refer to as "text." Her theory of "entextualisation," which builds on the anthropological and discursive theory of Greg Urban, is certainly important to create new corpora of African literature. My own interests here, however, are to assess how those corpora are actively kept apart by culturally loaded terms like "literature" (and, for that matter "folklore" or "oral traditions"). Indeed, it is precisely the definitional history of the literary, I argue, that has continued to produce such materials as nothing more than social "texts." It is for this reason that I use the term "literature," not to (yet again) subsume African expressive texts into a colonial ontology that they already labor under (as this paper will show), but, very much in thinking with Barber, to investigate the ways that the term has steadily created a regime of exclusion of African expressive texts and to imagine how future ontologies of literature and orature might undo such exclusions.

The discussion takes the form of a case study of the cataloging system at Makerere University Library. The reasons for my choice of Makerere University are patent, given the university's key position in debates and discussions about the politics of Anglophone African literature in the 1960s, as well as the fact that it was the training ground of major African literary scholars, activists, writers, and educators. This makes it the perfect context through which to think about the role of cataloging systems in the structuring of disciplinary and political knowledge. For this reason, the article focuses on (though is not restricted to) 1959 to 1964, which are crucial years in the history of African decolonization in the eastern African region." As Ngügi wa Thiong'o, whose autobiographical, dramatic, and novelistic writing will help me to frame and concretize my argument, puts it: "I entered Makerere University College in July 1959, subject of a British Crown Colony and left in March 1964, citizen of an independent African State" (Birth xi). These years at Makerere were significant to the field of African literature more broadly: in 1962, the university hosted the "Conference of African Writers of English Expression," of which Ngügi writes:

June 1962 came and I found myself among the big names of the time, which included Ezekiel Mphahlele, the main organizer, and Bloke Modisane, Lewis Nkosi, and Arthur Maimane-all South Africans in exile; Wole Soyinka, Chinua Achebe, Christopher Okigbo, J.P. Clark, and Donatus Nwoga, all from Nigeria; Kofi Awoonor . . . of Ghana; and our East African contingent of Grace Ogot, Rebecca Njau, and three Penpoint writers, Jonathan Kariara, John Nagenda, and me. (Birth 126)

This article investigates the relationship between the intellectual critiques of the colonial canon and curriculum that flourished at Makerere in the early 1960s and the institution of the library. In the heady urgency of political and literary debates at the time, libraries and the knowledge organization they performed and upheld were secondary to questions of representation of African literatures in the East African curriculum. Indeed, the library has remained backgrounded in discussions of the formation of the African literary canon, despite the fact that many of the standards that were operational in the mid-20th century continue to determine the field even to this day.

While the Makerere case study inevitably focuses on the classification systems at play during 1959 and 1964, it is important to note the continued relevance of the historical cataloging and classification standards I will be discussing in this paper. The physical library is seldom the site through which students are introduced to disciplines, or through which academic experts keep abreast with developments in their fields today. Most literary scholars rely on online search tools, like the World Catalogue, or even more routinely, Google, to search for primary and even scholarly materials. Yet, a key problem with searching data at the global scale that such search engines purport to operate at is that the sheer number of responses one gets for an inquiry produces the illusion of completeness. But it is important to note that any search tool is only as complete as the data it connects. As such, the problem is that the structural erasure of African literatures and oratures in previous knowledge organization systems are not corrected by these powerful search engines, but are, rather, reiterated and often redoubled by them-only, under the guise of producing even greater reach. This paper gestures toward what my larger study illustrates: that critically engaging the history of classification standards is crucial if we are to avoid passively accepting algorithms and AI technologies that blindly repeat colonial conceptions of the idea of "literature." What I aim to do here, instead, is to establish just how the library standards routinely introduced across Anglophone Africa in the mid-20th century have produced a problematically narrow ontology of the field of African literature.

In the service of this aim, this paper operates as part institutional history of the library at Makerere, part disciplinary history of the field of literary studies, part history of library classification systems, part personal history of Ngügi wa Thiong'o, and finally turns to a close-reading of Ngúgi's A Grain of Wheat (1967). This novel is set in 1963 at the moment of Kenya's independence and fictionalizes certain autobiographical experiences that Ngügi writes about in his memoir of his time at Makerere, Birth of a Dreamweaver. This close-reading argues that the library-as institution-which is not more than background to this novel and to the field of African literature more broadly, should, indeed, be foregrounded if we are to understand its impact on the creation of the field of African literature. This article thus foregrounds the story of the library as a colonial institution in order to illustrate its role in the making of the field of African literature.

MAKERERE UNIVERSITY LIBRARY

Makerere University hired its first librarian, a Miss Larter, in 1944 (Macpherson 41). By the 1950s, the library had outgrown its location in the main administrative building and, with the help of a Colonial Development and Welfare Grant, a new library building was erected in 1959. The new building was designed by British architects Nigel Norman and Graham Dawbarn, renowned for their "tropical modernism" in their design of the University College of the West Indies.· The newly minted library building in Makerere must have made quite an impression to the incoming student cohort of 1959, of which Ngúgi wa Thiong'o was а member. Unlike the iconic main university building, where the library had been located up until this point, and which Ngúgi wa Thiong'o claims was "a replica of the main building of the University College, London" (Birth 5),· the clean, subtly decorated, modernist architecture of the new library must have captured the bold spirit of the era of decolonization. Yet, unlike this new and decisive architecture, the history of Makerere library and its cataloging systems tell a tale that shows how the institution was deeply entangled in colonial classification and knowledge organization. What students of the 1959 cohort might have discovered on the shelves inside the library was by no means representative of the promise of the library's modern design.

University College London (UCL) was more than an emblem in the architecture of the main building at Makerere University College. By the time Ngügi started his studies, Makerere had already been an affiliate college of UCL for a decade, which had numerous implications for Makerere's curricula, examinations, and administration. The English department played a major role in the approval of the UCL affiliation in 1949, since a sticking point for the Inter-University Council Delegation, which held the power to approve Makerere's status as an affiliate of UCL, was incoming students' standard of English. Margaret MacPherson, who taught in the English department at Makerere from 1948 to 1981, outlines the final agreement that enabled the affiliation:

Not only must English teaching, and teaching in English, be improved both by fresh training and refresher courses, but the College must attack it in all faculties. All students must be required to pass a test of competence in English, while for the arts student taking a major course it was to be his classical and modern language as well as his study of literature. (60)

The English curriculum, which was thus central to Makerere University's broader education, required, like all education at the college, approval from University College London. We know from various accounts of the history of the English department that the curriculum in the late 1950s and early 1960s was still mostly the common fare of high British canonical works (Dipio 78), as hinted at in the tradition of the production of an "annual Shakespeare play for first-year arts students" (Macpherson 173). Ngiigi himself is more direct about this content, noting "the Leavisite moral scheme of 'high' versus low" culture that were [sic.] prevalent in the Makerere English Department" (Birth 185). He reminisces that the "writers [he had] read for [his] English classes" included "Charles Dickens, Jane Austen, Emily Bronté, George Eliot, Thomas Hardy, D. H. Lawrence," as well as Joseph Conrad [and] Graham Greene" (Birth 84, 106). In 1943, a year before the appointment of Makerere's first librarian (when acquisitions were largely driven by the needs created by course content), Margaret Macpherson reports that a "grant from the Carnegie Corporation of £1350 enabled a start to be made in the collection of the much-needed standard texts that the College had not yet been able to acquire" (39). When it comes to literary studies, we can imagine that these standard works were very much those purposed for the English department's literary, and canonical, curriculum.

Alongside this "standard" content of the library, I want to pay attention to another set of standards that organized that content and choreographed students' exploration of the books in the university's library. The library at Makerere University adopted the Dewey Decimal Classification system in 1944, the same year as Miss Larter was appointed (MacPherson 41). One of Miss Larter's main ambitions on her appointment, which would be followed up by successive head librarians, was to make Makerere Library the formal archive for public records from the entire East African region.' This ambition is relevant to my discussion because in its service the library needed to demonstrate its professionalism and reliability through its careful adherence to top international standards. We see this in a letter exchange between the archivist of the Central Government Offices of the Chief Secretary of Kenya and the head librarian, Miss Larter's successor, in 1956. The unnamed Kenyan archivist writes: "Presumably your library books are being classified strictly in accordance with the Dewey Universal Decimal System," to which the librarian reassuringly responds that the library, indeed, classifies its "books in accordance with the Dewey Decimal Classification." The expressed presumption on behalf of the Kenyan archivist-and the tone with which it is delivered-illustrates the extent to which the use of Dewey Decimal Classification standard (hereafter DDC) assured best practice at that time. Any serious library, especially one with hopes of becoming a regional archive, was required to have adopted the DDC.

The DDC was, and remains even today, one of the world's most commonly used systems for shelving numbers in libraries. Before the digitization of library browsing, most literary scholars, for example, knew to go directly to the 800category (Literature) in a library using the DDC. Thomas Mann, who was a consultant in Germanic languages at the Library of Congress, wrote in "On the Essential Importance of Collections of Books Shelved in Subject Classified Arrangements" that the

majority of faculty and students recognize from their own direct experience that focused depth searching of the contents of most of the copyrighted books on a particular topic cannot be realistically done in any way other than the systematic browsing of subject-classified bookstacks [and] serendipitous discovery by recognition-browsing within carefully defined segments of library book collections is crucial to many research projects because it enables researchers to find relevant sources whose keywords they cannot specify in advance. (Mann 199; qtd. in Green and Rathbun-Grubb 271)

While most researchers may use online catalogs or search engines instead of physically browsing the library today the problem Mann outlines remains: how researchers browse digitally is, much like the physical shelves of a library, largely determined by an underlying data structure-in this case, the DDC. Indeed, the DDC provided just such a subject classification system for shelves in libraries. As such the metadata captured in the DDC system was far more than an abstract and neutral numerical code. Instead, it maneuvered library users' exploration of the library and directed their corporeal experience of searching (and thus researching) the library. Without going too deeply into the history of the DDC, what is relevant for this discussion is the fact that the system is based on ten classes each housing ten sub-classes, with the possibility of endless decimal expansion. This is what the 800- and 890-ranges looked like in the 1958 edition of DDC, which would have been implemented for use at Makerere Library in the period under discussion. Apart from an update to inaccurate and politically problematic language, the DDC remains structurally largely unchanged to this day."

The problem of this model for African literature is immediately apparent when we see the imposed hierarchies on the literary field (American and English on top, followed by European), as well as the prioritization of European language- literatures at a higher level of the ontology (the 800-range) while the rest of the world's literatures are shunted into the sub-range of 890, "Other Literatures." This impacted visibility and coherence of African literature in very concrete ways: "Other Literatures" were placed furthest away from literatures of the Western world; the categories of the 890-range were less likely to appear on coordinating data on library shelves; the sub-range of "Other Literatures" is presented as having some invisible binding principle behind it, and-perhaps most consequentially- literature written by Africans in English, French, or Portuguese became separated from African language literatures.

This also points to a further ontological problem underlying this structure: that being what was considered self-evidently part of the category of "literature" in the first place. Eunice Kua has pointed to the all-important conceptual split in the DDC between "Literature" (under 800) and "Folk Literature" or "Folklore" (under 398), arguing that the "separation of literary work and folklore potentially leads to... fragmentation, especially in cultures such as those of Africa, where there is a strong oral tradition, and the boundaries between literature and folk literature are blurred" (259). This is a problem shared by all classifications of literature in systems devised in the 19th and 20th centuries. If we take a look at the hierarchy in the above table, we see that the sub-ranges 896-899 are all literary traditions that include a strong oral history, hinting toward the ontological challenge that they posed to the very concept of literature.

In the DDC, the 398 range covers "Folkore," which is a sub-range under Social Sciences (390), rather than Literature (800): this in itself indicates the extent to which non-European literary and oratory cultures were considered to be an appendix to ethnography, rather than having aesthetic value in themselves. It is important to note that this split has never been addressed in revisions of the DDC, which leads to clear cultural bias and barely concealed racism even in today's 23rd edition. For example, librarians are exhorted in DDC23 to "Class an anonymous classic or a work by a known author with the classic or work in 800, e.g., Icelandic sagas 839, fairy tales by Hans Christian Andersen 839.81" (DDC23 300 551), that is to say, under the "Literature" range. This is interesting, given that sagas and fairytales do not fit in the express genres listed under "Literature (Belles Lettres)" (those being "Drama," "Poetry," "Epigrams," "Fiction," "Essays," "Speeches," "Letters," "Miscellany," and "Humor and Satire", DDC23 800, 879). While the system seems clear about instructing librarians to "class folk literature in 398.2" (DDC23 800, 879), riddles at 398.6, and proverbs at 398.9 (DDC23 300, 551), the question emerges as to whether the formal and generic differences between these folkloric forms and the stories of Hans Christian Andersen is not at least partly motivated by the ethnicities of the writers.

The fact that the DDC recommends cross-referencing folktales to ethnic groups bears out this argument. The range "Ethnic and National Groups" is itself a problematic and relatively new range in the DDC. It was created in the 2003 edition to replace the classification "basic races." In the 1996, DDC21, the range "basic races" included "caucasoids, mongoloids, and negroids," as well as "mixtures of basic races" (from table 5 "Racial, Ethnic, National Groups," DDC 21st edition 1996, qtd. in Furner 156). The editor of the DDC22, Joan Mitchell, indicated that this change was made "to reflect the de-emphasis on race in current scholarship" (Mitchell 28).

This is what the range "ethnic and national groups", which replaced "basic races," and which we find, like "Folklore," under the 300-range, looks like since DDC22:

The first thing that strikes a close-reader of these updated tables is that this structure mirrors the 400-ranges for Languages (see DDC 400) and the 800-ranges for Literature (see Figure 1 above), which possibly is justified by the move away from biological terms of "race" toward cultural terms of "ethnicity." However, as Jonathan Furner has argued in "Dewey Deracialized: A Critical Race-Theoretical Perspective," the problem here is that the decision to "deracialize the DDC" has, inadvertently, led to a "color-blindness that is encouraged by the liberal racetheoretic approach," which has "the effect merely of sustaining the hegemonic status quo in which discrimination and economic and social inequities in favor of whites are institutionally maintained" (Furner 164). The hierarchical relationship in the table above of European groups to "other ethnic groups"-and the definitional line drawn between the two-is a case in point. Further to that, I argue that we can see that by turning earlier language and literary classification ranges and their hierarchies into the basis of definitions of people, DDC has imported the cultural biases of the period in which it was devised (1876) into a contemporary model of understanding ethnic and national groups today. The cross-referencing of folklore (but not literature) to such ethnic classes, then, reveals a colonial attitude to non-white culture, orature, or literature. Or to put it more simply, it is unlikely that one would find the Icelandic sagas and Hans Christian Andersen's tales crossreferenced to an ethnic code in the DDC.

When it comes to the category of African folk literatures, today's DDC23 does give a nod toward "Folk Poetry" and "Folk Drama" being classified under rhetoric, which is nested under "Literature" (808.81 and 808.82, respectively) (DDC23 300, 551), though this is a relatively new development and would not have been the case in Makerere during the time of my case study. That said, one sees evidence of a similar porosity between folk literatures and belles lettres in today's catalog at Makerere, though it is impossible to know when these specific interventions were made. This porosity is specifically apparent where texts categorized under the subject heading "Folk literatures" are variously cataloged at 398 "Folklore," 808 "Rhetoric & Collections of Literature," and 896 "African Literature" (of which, more below). To consider the effect of this for, for example, Kenyan orature, we can see that Wanjiku Mukabi Kabira's Gikuyu Oral Literatures (Heinemann, 1988) is classified under "Folklore," while his book with Karega wa Mutahi, Kenyan Oral Narratives: A Selection, is classified at 808, and Christon 5. Mwakasaka's The Oral Literature of the Banyakyusa (Kenyan Literature Bureau, 1978) earns a classification under 896, "African Literature." The consequence is a highly fragmented and dispersed "field" of Kenyan oral literatures.

We can glean two points from this that are key to my argumentation here: first, the DDC was, and is still, shaped by the colonial and book-based knowledge context out of which it emerged; and second, this has had particularly negative consequences for findability and visibility on library shelves of African literatures that are either written in African languages or are oral or performed. If we consider the extent to which oral and performed forms were central to the politics of the literary curriculum debates in East Africa in the 1960s'·-a centrality reinforced by Ngügi wa Thiong'o's argument about these forms in his later Decolonizing the Mind: The Politics of Language in African Literature-we start to see that cataloging practices have been far slower than curriculum reform to respond to the politics of decolonization. Indeed, the fact that even these more recent editions of the DDC continue to entrench the colonial epistemes of the system that determined the spatial logic of Makerere library in the early 1960s illustrates just how inert the standard has been in its response to real political change.

This sluggish response of catalog systems to political change is precisely because of their purported global reach and scope, their complexity, and their historical hold on the ontologies of knowledge often taken for granted. Cataloging structures and systems are backgrounded because outside of library science they have been seen as neutral and merely descriptive aids for finding materials. Also, to understand the part of the system that is relevant to any given discipline, one needs to understand its relation to the whole. Although resistant to transformation, changes in the use of these systems by librarians can be traced in local interpretations of cataloging systems. We see an example of this in Makerere Library's current MARC records," where a local intervention has been made to address the problem of the invisibility of African literature on the library's shelves. From the first DDC to the present, the range 827 denotes "English Humor and Satire" (DDC23 Summaries 17). At Makerere this code has been extended to cover 20thcentury English-language fiction, with African authored literatures in English being collected under 827.9669. In the MARC records, this is recorded a "locally assigned DDC number" (092 in MARC).

The local reassignment of a DDC field that is not relevant to this particular library ("English Humor and Satire") to a body of work requiring greater visibility makes a lot of sense at the local library level. What is striking, however, about the creation of a library-specific DDC code for African literature is the fact that the DDC has actually had a category for African Literatures since 1876 (see table 896 in table above)" However, it is clear that, because the 896 code is nested under "Literatures of other languages," there has been reticence in placing Englishlanguage African literature under this range even at Makerere's library. The outcome of this is the entrenchment of the fundamental split in the field, between English or French African literatures and those in indigenous languages that 1 mentioned before.

Various and inconsistent linguistic and geographical motivations for including materials in certain categories across the DDC compound this matter further. We see such inconsistency in Makerere University Library today: the "African literature" (896) range, typically set aside for African language literature, actually includes 357 English-language texts. Whether these English texts were assigned to the 896 range before Makerere decided to use the 827 range ("English Humor and Satire") for Anglophone African literature, or whether this indicates the beginning of a multilingual category in the library, is unclear, and I have not been able to get a response to this question from the current cataloger. Whatever the reasons, there are confusions across the classification scheme when it comes to language and region. For example, the English-language fiction of Ugandan writer Goretti Kyomuhendo is dispersed between 8% ("African Literature"), 827 (Makerere's local code for "African Literature in English"), and 823 ("English Fiction"). When it comes to the African-language literature code (896), it is notable that relatively few African language literatures are placed here (thirty-nine Swahili, seven Ganda, three Nyoro, and two Nyankole) and even fewer texts in European languages other than English (nine French, three German, one Greek, and one Dutch).

What all of this evidences is a lack of clarity as to the space of African literary works in the catalog. African literature floods the barriers that attempt to contain the ontologies of the literary. Concretely, what this means for the library shelf, is that even today, the works of a writer like Ngúgi wa Thiong'o would be split on the shelf between his English-language materials (827) and his Gikuyu texts (896): a particularly harsh irony given Ngiigi's politics on this point. However, since the library at Makerere does not have any of Ngügi's works in Gikuyu in its collections-raising the further problem of acquisitions of African-language content in libraries across the continent-this is a moot point.

If we return to what the student cohort of 1959 might have experienced when browsing the shelves of their newly built library, we can imagine that the combination of the donations of "standard works," the English literature curriculum that was so pervasive across Makerere's courses, and the use of DDC call numbers (either of the 15th edition of 1951 or the 16th of 1958) for shelving this content, would have made for a full immersion in the ontologies and content of the colonial library. In Birth of a Dreamweaver: A Writer's Awakening (2016), Ngügi wa Thiong'o writes: "Makerere taught me to value books . . . ; her well-endowed library became my second residence. In the library, I was the lord of the intellectual manor with a hierarchy of willing and dedicated staff ready to serve me" (65). He then writes one of the least presumptuous sentences in the book: "That was how I stumbled upon Norman Leys's Kenya." (65) According to Ngügi's account, Leys's Kenya plays a significant role in his "awakening," not only as a writer, but as a political writer: Leys was a welcome alternative to the colonial commonplaces of Elspeth Huxley that had so tired Ngúgi by this point. But it is the verb "stumbled" that warrants focus here. Ngügi was not aware of Keys until "stumbling upon" his work in the Makerere library. The verb suggests happenstance, an element of randomness that is evoked by the experience of browsing. But, as I have been arguing, the workings of knowledge organization in any library are much more profoundly choreographed than that.

What anyone might have "stumbled upon" in Makerere library at this time was first and foremost likely to be within the range of call numbers, in this case the Dewey Decimal numbers, that determined the placement of the book on the shelves. But the card catalog also played an important role in shaping what library users might happen upon. In Makerere's case, the use of the Library of Congress subject headings with DDC call numbers allowed for at least some porousness of categories in ways that would have allowed for more spontaneous discovery of African written materials. If we were to speculate on how Ngügi happened upon Leys's Kenya in the Makerere library, we can imagine him either browsing the "History, Africa" section of the DDC organized shelf, where the book was classified; or flipping through the card catalog under the subject heading "Kenya." Such subject-headings, which would have been newly implemented during Ngügi's student years at Makerere, are even now the key coordinates of the analog card catalog at Makerere today, where, despite the fact that the full catalog is now available online, students are still required to learn how to use the card catalog because of the regularity of power-outages and the high costs of mobile data in Kampala."

THE CARD CATALOG AND THE LIBRARY OF CONGRESS SUBJECT HEADINGS

During the 20th century, the system used almost universally for the organization and discovery of materials in libraries was the card-catalog system, a system that had been nationalized and then internationalized by the Library of Congress's card standardization and distribution program from 1901 onward (Edlund 384). The Library of Congress's card distribution system was created to save librarians across the United States the trouble of writing out catalog cards for every new item that came into their collections. By preprinting and making available for order cards that described books in their considerable collection, smaller libraries across the country had most of their needs serviced by this scheme. Over time, this became a global system, and we know from a Manual of Cataloging Procedures published by Makerere library in 1972 that the library had, at that point, standing orders for Library of Congress (hereafter LC) cards. If one browses through the physical card catalog at Makerere one comes across many examples of LC cards. The cards made for order were author cards, meaning that they left coordinating metadata free so that librarians could add their own shelving systems accordingly. A basic LC author card would include the author surname and name, followed by the title and publication data, as well as a basic description of the material dimensions of the item.

This purely descriptive, and thus putatively neutral, information on LC cards was, however, accompanied by a further coordinating piece of metadata: a subjectheading. In the circular announcing the card distribution project in 1901, we read that the Library of Congress "prints by way of memorandum on the card the subject-headings that it will use on the copies destined for subject cards" (Edlund 384). This is the crucial point at which the descriptive metadata of the author card gets enmeshed in an ontology, or knowledge hierarchy, since the subject-headings are prescriptive and thus culturally, historically, and ideologically informed. At Makerere University Library, where the DDC system structures the manifestation of its categories on the shelf, the LC subject-headings are most noticeable in how they coordinate the experience of browsing a subject in the card catalog.'

The library at Makerere already saw the need for a subject index in the 1950s. The increasing size of the library was a problem addressed in Makerere library's 1952 report where the main librarian, Miss Larter, hinted at the future need for a subject index, if only in passing." By 1956, the library had grown so quickly that this need was more pressing. By then the library had prepared a report "on the need for and the cost of a Subject Index to the Library's collections." It is not entirely clear when the Library of Congress's subject headings were introduced, but we can assume that the process began shortly after the 1956 report. We can be sure, at least, that by the end of the 1950s and into the early 1960s the Library of Congress subject headings were used to organize the library's subject catalog.

At the same time as Makerere was implementing the LC subject headings, the term "African Literature" appeared for the first time in the July-August 1960 supplement to the 6th edition of the LC subject headings (hereafter LCSH),"® which meant that the term was included in the 7th edition of the LC Subject Catalog in 1966 (see Quattlebaum 23). That the field of African literature becomes authorized in the Subject Catalog in 1960 is not entirely surprising given that year's relevance to decolonization in Africa. This political history is also reflected in the fact that the Library of Congress created an "African Section in the General Reference and Bibliography Division in 1960" (Witherall 127). Also, just two years before, African literature had definitively entered world literature via the launch of the Heinemann African Writers' series in 1958. During the time of Makerere library's employment of the LCSH, then, the field of African literature was changing globally.

Another event that impacted African literature's place on the global map around this time was the aforementioned "Conference of African Writers of English Expression" of 1962 at Makerere University. Ngúgi wa Thiong'o points out that one of the key concerns of that conference was a "preoccupation with definition" (Decolonising 22) of African literature. For Ngiigi, this preoccupation betrayed a "crisis of identity" of the African petty-bourgeoisie: "The literature [the petty-bourgeoise] produced in European languages was given the identity of African literature as if there had never been literature in African languages," he writes (Decolonising 22). The statement echoes my argument above about the fragmentation of the field produced by the DDC and if we take a close look at the history of the LC subject catalog, this takes on even greater empirical weight. While a few African language literatures had been represented in previous editions of the LC subject headings under various languages (most notably, Swahili Literature), the creation of an African Literature category in the 1960s seemed first and foremost to address the significant increase in African book publication in European languages at the time of decolonization. The creation of the category of African literature, then, might speak more to the creation of a putative split between European-language African literature and African-language literature than to the sanction of African language literatures in world literary studies per se.

The fragmentation this causes, as for that observed above in the DDC, was discussed by Robert L. Mowery in a 1973 article titled "The Classification of African Literature by the Library of Congress."" Mowery writes: "Neither the Library of Congress (LC) Classification nor the Dewey Decimal Classification provides a single classification range for African literature. Since both systems classify literary works by the languages in which they were originally written, they scatter the various literatures of Africa throughout the schedules for English literature, French literature, Swahili literature, Portuguese literature, and various other literatures" (340). However, as Mowery's conflation of all African languages, other than Swahili, into the somewhat dismissive "and various other languages" above suggests, his paper is primarily concerned with creating a coherent solution for Anglophone African literatures. Indeed, the problem of African literature emerges for him as a practical concern at his own work place (Illinois Wesleyan University library) because of the

problem of classifying the initial eighty-three titles of the African Writers Series (AWS) published during the past decade by Heinemann Educational Books, Ltd. Composed of primary works written in English or translated into English from Arabic, French, Portuguese, or various African languages, this respected series contains works of individual authors as well as collections and anthologies. Even though poetry, plays, essays, short stories, and novels appear in this series, short stories and novels predominate. (Mowery 346)

Mowery's method speaks to the entanglements of the book and publishing industry with library acquisitions and metadata systems and also provides some backing to my earlier comment that African literature enters the LC catalog in part because of the Heinemann African Writers' Series' successes at this point. The problem here, however, is that the body of work that the LC addressed in creating its "African Literature" subject-heading-as important as it has been for the canonization of African writing as world literature-has produced African literature as a predominantly European-language literature. Mowery's key concern, then, is to improve the "success of finding the works of English-language African writers" (349), and he makes recommendations to this effect. In a postscript to his article, he notes that the LC Subject Cataloguing division has "not only accepted the principle of [his] proposal but also has incorporated it within a recently-published revision of the schedules for "English literature outside of Great Britain" (Mowery 351). That the creative works in the English language should be cataloged on the periphery of an implied central body of British English works is obviously problematic and does not address the problem I am trying to articulate here: that being, how global catalog systems fragment and occlude African literatures, and-as we see in the solution adopted by the LC in 1973·-assimilate African writings into colonial language canons. Following Michel Foucault's argument in The Archeology of Knowledge I see such displacements of the field of African literature as Foucauldian strategies (64) of colonial knowledge organization, which is also to say that colonial ideology is both traceable through, and consolidated by, book metadata structures such as the LC subject catalog.

THE SUBJECT OF AFRICAN LITERATURE

Before "African Literature" entered the subject-catalog of the LC one would have either found African-authored texts in European languages classified under the English (or other European language) literature subject cards. We can still see the remnants of this organizational structure under today's subject, "African literature," which unlike "English Literature" or "Folk Literature" is not subdivided by geographical region. Instead, the current "African Literature" category is subdivided into European language categories (English, French, Portuguese and Spanish), suggesting that literary works from the continent are extensions of European, rather than African, literary traditions.

Even in the current LC subject-catalog "Literature" is to be used for "Belleslettres," "Western literature (Western countries)," and "World literature" (LCSH, L 244), thus driving a classificatory divide between Western and world literature, which itself betrays a locational bias. We are then instructed to find individual language literatures under "headings of the type [language] literature, e.g., Swahili literature; also headings for national literatures, e.g., American literature; Mexican literature" (LCSH, L 244; square brackets in the original). As such, the LC subject catalog variously and inconsistently uses two classificatory principles to classify literature: language and region. This confusion allows for cultural bias and different cultural valuations of the categories at play to enter the ontology. For example, English Literature is not a broader term for American Literature (LCSH, A 233), yet the different language varieties included within American literature do not expressly include "English language literature," since English is taken as the given language of American literature (LCSH, A 235). As such the linguistic principle and the regional one are collapsed without comment.

Furthermore, the "English Literature" class includes national literatures from across Africa written in English, such as "Kenyan Literature (English)" and "Ugandan Literature (English) (LCSH, E 183), which would separate the field of "Kenyan Literature (English)" from all the other African language literatures in Kenya. The field "Kenya-Literatures" (LCSH, K 72) does include African Kenyan languages (interestingly Swahili is not noted), but "Kenyan Literature (English)" and "Kenyan Literature (French)" are also represented at the same hierarchical level as "Kenyan Literature," suggesting their greater visibility in the scheme than Kenyan literatures in African languages. Things get even more complicated if we look at the literary genres of "Fiction," "Poetry," and "Drama," as listed under the national determiner "Kenya." In these schedules only English- or French-language texts are included. Thus, if you search for a Kenyan poem in a language other than English (or, oddly, French) under "Kenyan Poetry," you would find nothing. For that you would have to search under the poetry sub-categories of "Swahili," "Kikuyu" [sic], and other Kenyan language literatures.

Of course, this is not only a politics of language but also of materiality: we might well ask ourselves what the LCSH does with non-book literatures. "Oral Tradition" (LCSH, O 109) is subordinate to "Oral Communication" and is noted as a "relative term" (RT) to "Folklore." The term "Oratory" has replaced the more ideologically revealing "Oratory, Primitive [Former heading]" (LCSH, O 111). But, in the current version, "Oratory" is almost entirely understood as being within the field of rhetoric and does not cover the literary expression of African oral forms at all. "Folklore" is also to be "used for" (UF) "Folk Beliefs" (LCSH, F 185), which indicates the ethnographic understanding underlying this category. The subcategories of "Oral Tradition," "Storytelling," and "Folk Literature" (LCSH, F 184-85) are then all conceptually linked to an ethnographic understanding of culture, which explains why "Folklore" can also be correlated with regions like "Folklore- Africa, East," or national geographies, like "Folklore-Congo (Democratic Republic)," which would then include a wider variety of different kinds of folkloric arts.

Indeed, "Folk Literature"-which gets closest to the kinds of orature and literature I am interested in-is listed in the same sweep as "Folk Arts," "Folk Dance," "Folk Drama," "Folk Poetry," and "Folk Songs," begging the question how these categories differ from the simple categories of "Dance," "Drama," "Poetry," and "Songs." "Folk Literature" includes among its sub-categories the following genres or material forms: "Chapbooks," "Folk drama," "Folk poetry," "Folk songs," "Proverbs," "Riddles," "Tales," "Tongue twisters" (LCSH, F 171). It is interesting that both written and oral forms are collapsed under the heading "Folk Literature," but the promise of this blurring of lines between the written and the oral still betrays a Western bias (and is not included under the "Literature" subject heading). The written forms here are those without cultural capital (chapbooks, not novels), and similarly, "Folk Poetry" seems not to include "Laudatory Poetry" and "Epic Poetry," two of the most important oral genres across the African continent, since these have their own Subject Headings (which are, in the first instance, universal and are then narrowed according to language or region). This once again scatters the body of expressive literatures and oratures of Africa across the subject catalog.

A further observation on the hierarchical idiosyncrasy of the LCSH for African literature is that both "Folklore" and, since 1966, "African Literature" are higher order categories for "Folk Literature", yet "Folklore" (regionally attached to the African continent) and "African Literature" are not considered relative terms of one another, which would at least enable movement between the categories. Relative terms for "Folklore" do, however, include "Material Culture," "Oral Tradition," and "Storytelling" (LCSH, F 184). What we see here, then, is a clear ontological split between the forms considered African (nested under "Folklore") and those that carry with them the ontological weight of being Western (books written in European languages). More concretely, we can see in the hierarchy of today's subject-headings evidence of the conceptual split between the oral and literary in the field of African literature, a split that has large consequences for African language literatures, too. These idiosyncrasies are legacies of the colonial history of bibliographic classification and cataloging. This does not mean, of course, that the LCSH has not undergone important changes in the wake of political change over time." Yet, while various political advancements have led to change, these are not by-and-large structural and do not solve the problem of a fragmented and obscure archive when it comes to African literary works.

We can only surmise how rigid the card-catalog would have been during Ngügi's time at Makerere, when little published African literature (even in European languages) was in circulation and even less collected in the university's nascent collections. Ngúgi would go on to articulate the problem at the level of the literary curriculum in his 1968 manifesto with Taban Lo Liyong and Henry OwuorAnyumba, "On the Abolition of the English Department." I am, here, extending their argument to cover the institution of the library and the strategic discourses and concrete systems and schemas that order its work: what Ngúgi would later claim was the "boldest" part of "On the Abolition of the English Department" was the call "for the placing, within the national perspective, of oral literature (orature) at the centre of the syllabus" (Decolonising 94). The question is, then, what happens to the category of "African Literature" in the cataloging systems of the world if we were to put "orature at the centre," or at least give it equivalent status to written forms, in the conception of the field of African literature? Perhaps one place to start would be at the level of knowledge organization and ontology in the descriptive catalog that have, I am arguing, de facto consequences for conceptions and understandings of the field.

In his memoir Birth of a Dreamweaver, Ngúgi's analysis does not extend to the ways in which the library catalog of his time divided what he learned in his literary studies and the literary knowledge he learned in his

mother's hut where [he] first heard her stories of Swallow, who carried messages of those in distress; Hare, who fooled the biggies, Hyena, Lion, and Leopard, even arbitrating among them; Donkey, who brayed sorrow and shat mountains at the same time but also was so stubborn; and the scary two-mouthed ogre who lured beautiful maidens into his lair up in baobab and sycamore trees. (75)

He does, however, articulate his frustrations in terms directly pertinent to a major aspect of the classification and cataloging systems in wide use during his time at Makerere. He writes, in Birth of a Dreamweaver, that the

settlers' claims that Kenya was the White Highlands. . . had been given literary immortality in the work of Elspeth Huxley. The colonial state sided with the literary, but the oral was the voice of a people, and even after the state banned FLA songs and poetry, song and dance inspired defiance in those herded in barbed-wired trucks into concentration camps. (103)

The literary, then, clearly the realm of English department curricula, of settler culture, and of the novelistic tradition, is set up antithetically to the "voice of a people," the oral traditions that the library would classify under "Folk Literature."

Indeed, questions of classification were clearly on the minds of the young literary Makerereans at the time. The East African Literature Bureau ran a "novelwriting competition in 1961 open to all East Africans with a prize of one thousand shillings" financed by the Rockefeller Foundation (Ngügi, Birth 83). Ngúgi and his friend Joe Mitiga decided to enter the competition, and when Joe Mútiga read ŞA

Ngúgi "bits and pieces of what веет [ед] to be a folktale," Ngügi "raise[d] concerns," only to be told by Miitiga that "they [the bureau] didn't say that a fable couldn't be a novel" (Birth 86). Joe Mútiga would never complete his fable-novel, which is a pity since, among other things, it would have been valuable to see how such a text would have been categorized in the Makerere catalog. Ngügi's entry to the competition would go on to be published as The River Between, a book he was ambivalent about. On the one hand, he celebrated the fact, "I have dared to do something that has not been done in Kenya or East Africa: write a novel" (Birth 98). But he was also embarrassed at his investment in "following in the footsteps" of the canonical writers he was reading for his English class (84). "Heinemann," he writes, "had published some the canonical authors I was reading in my honours class: Joseph Conrad, Graham Greene, D. H. Lawrence. Was I going to join the club? Could this be happening to me?" (Birth 160).

Wary of becoming a member of this club (the "petty-bourgeoise" that he derided), it is not insignificant that Ngügi returned to writing drama after his first forays into the novel form. His turn to drama in The Black Hermit, while distinctly influenced by the Shakespearean tradition he had been taught at Makerere, nevertheless suggests the inadequacy of the novel form for his creative expression and thus preempts his ideas about drama as a crucial genre of African literature in Decolonising the Mind. Although the play was first published by Makerere University Press in 1963, the Makerere Library catalog only lists its Heinemann 1968 and 1981 editions today," where one version is cataloged under the DDC local category for African literature (827) and the other as English Drama (822)-a split that is not without scholarly irony, given Ngügi's argument in Decolonising the Mind, but that also reveals the inheritance of the literary classifications that go largely unnoticed in library structures to this day.

THE LIBRARY AS BACKGROUND IN A GRAIN OF WHEAT

If we close read Ngúgi for library metadata we come up short; but the library as an institution of colonial knowledge does figure as background in his novel, A Grain of Wheat (1967). I approach the depiction of the library in Grain of Wheat with Louise Green's productive method of "reading for background," which binds an analysis of the "minute detail and the particular work of description" (26) that is "not in any way relevant to the plot" (27) of a novel. Such "background" description "gives weight to the narrative through anchoring it in the material world," writes Green (27), but in analyzing such background, her method also encourages us to pay close and refined attention to the ways in which the material world shapes and limits the action of the novel under analysis. While Green puts this method to work on environmental concerns, I find it generative in reading the library as background to Ngiigi's historical novel, where Kenyan independence is foregrounded and is critical to the plot of the novel, as the action leads up to the Uhuru celebrations. Amid this major historical plot of Kenyan independence, it would be easy to miss the library altogether or to dismiss it as simply background detail.

Yet, the library in A Grain of Wheat is one of the autobiographical elements of the novel, partly based on Ngügi's time working at a research library for the East African Agricultural and Forestry Research Organization in Kenya. In Birth of a Dreamweaver Ngúgi depicts this experience, which took place "early in his first year" of his studies (106), as part of his political consciousness raising. It is not clear from his memoir what his own role at the library was, but this role is less significant than that of Moses Wainaina (not his real name), the "African assistant" of the head of the library division (Birth 106) who is recreated as Karanja, one of the most important characters in the plot of A Grain of Wheat. Karanja carries the themes of betrayal and colonial assimilation throughout A Grain of Wheat, and the depiction of Moses Wainaina in Birth of a Dreamweaver similarly suggests a man who was a toady to white power and who abused his own position of power in the library by brutalizing those below him in the institution's hierarchy. The character Karanja, too, enjoys the power that he is given as a "homeguard," acting on behalf of the British to quell the Land and Freedom Army (LFA) (143). To do this, Karanja has to go back on the oath he himself had taken with the LFA. That he is, in the present of the novel, a librarian for a colonial research organization for Agricultural and Forestry is a wry comment on the extent to which he has sold the ideals of the LFA for his own personal gain.

While Karanja is ultimately not the betrayer of the revolutionary Kihika, whose absence the novel circulates around, his lack of political conviction, his servility to, and protection by, white colonists, and his protection of that power through small and large cruelties in the novel, all speak to a larger, if more banal, betrayal. Unlike the protagonist Mugo, who admits his betrayal of Kihika to the entire community, Karanja lacks the moral courage to admit any of his wrongs and this makes his betrayal all the worse. Ngiigi allows the novel structure to highlight how this perpetual, everyday level of betrayal accumulates into something worse than Mugo's moment of treachery because, unlike Mugo, Karanja does not feel guilty about his actions: they simply fade into the background.

Karanja's library work is complexly tied to colonial power systems. The following passage stitches the prosaic work of writing labels for books to both the land struggle of the LFA and to Karanja's fear of Uhuru (freedom) and his active repression of the fact that it is impending:

Karanja picked a clean stencil from a pile on the table and started writing labels. The books recently bound at Githima belonged to the Ministry of Agriculture, Nairobi. Soon Karanja's mind lost consciousness of other things. Uhuru and Dr Van Dyke, and he concentrated on the label in hand: STUDIES IN AGRONOMY VOL.-. (34)

His pedantic attention to detail in his work "at Githima Library dusting books, keeping them straight in their shelves and writing labels . . ." (34), like his "obsequious attention" (35) to his white boss, is in the service of maintaining the "good name he had built up for himself among the white people" (36). His investment in that "good name" is because of his buy-in to a hierarchy of colonial power. Later we read that when he was a homeguard and then a Chief, killing had made him feel "a part of an invisible might whose symbol was the whiteman [sic.]. Later, this consciousness of power, this ability to dispose of human life by merely pulling a trigger, so obsessed him that it became a need." (230) His abuse of power comes out in how he treats the black workers under his authority at the library. When one worker interrupts him while he is writing a label he is enraged, complaining that "some people don't understand that the work we do, you know, writing labels for all those books of science, requires concentration" (40). Karanja implies that "some people," a phrase delicately developed in the novel as his deprecating term for black people, do not understand "science," articulated discursively as a field of white rationality.

Of course, Karanja has no real power in this colonial world and most of the action in the present of the novel revolves around his somewhat pathetic attempts to pick up enough courage to simply ask his boss, John Thompson, if he is leaving Kenya in advance of Uhuru. Thompson is, not insignificantly, having an affair with the main Librarian, Mrs Dickinson; a background fact that nevertheless highlights Karanja's servility as he becomes little more than a messenger between the illicit pair. Thompson also focalizes an important scene that occurs just outside the library. In the scene, seen through the eyes of Thompson from his office window, a dog of one of the researchers at the research institute goes to attack Karanja and he picks up a stone to protect himself (Grain 42). Though he does not throw the stone, the scene is a fictional microcosm of a much bloodier event in Kenyan history. In Birth of a Dreamweaver Ngúgi writes about "the day a dog, a pistol, and a stone collided in one Nairobi street":

The dog and the pistol belonged to Peter Pool, a soldier who had fought against the LFA and then set up an engineering shop on Government Road, Nairobi, now Moi Road. The stone belonged to Kamawe Musunge, Poole's houseworker, then called a houseboy. For reasons unknown, maybe for the usual sport, Peter Poole set two dogs on Musunge, who was riding a bicycle. We know that Poole was the aggressor, for why else would the dogs attack, on their own, a person they had often seen in and about the house? Musunge, frightened, picked up a stone for self-defense. He didn't throw the stone, but Poole shot him dead for threatening his two attack dogs. (70-71)

The event in A Grain of Wheat is far less bloody: Karanja is not harmed as a result of threatening the aggressive dog, even though the dog's owner bays for blood. But the incident is important to illustrate the relevance of the library as background. The scene could be read as though staged, in which the library is, quite literally, the backdrop of the action. The action itself is a small and banal event against the backdrop of Kenya's independence, but Ngúgi masterfully insists we pause and pay attention to colonial power at this banal level by using the scene to replot the true story of Peter Poole and Kamawe Musunge. As such the backdrop of the library, the seeming nonevent of the scene, becomes stitched to-and indicative of-the massive historical implications of white colonial power in East Africa. It also sets up the full impact of the following line: after Karanja has finally heard from Thompson that he is leaving Kenya, he returns to the library, "his gait, to an observer, conjuring up the picture of a dog that has been unexpectedly snubbed by the master it trusts" (161). As such, the library can be read as a synecdoche for the epistemic and classificatory violences it encodes. Karanja's sycophancy is, if we push the metonymy further, not unlike what Ngúgi sees as the petty bourgeois ambitions of the writers at the 1962 Makerere conference. By staging this scene of barely contained physical violence in front of the library, I argue that the novel might be understood to acknowledge its own complicity in the very system it wishes to critique, its own wish to "join the club" of esteemed canonical authors (Birth 160) while simultaneously "making" a new, African, literature. Ngúgi would, of course, go on to articulate this complicity of Anglophone African literature a year later in "On the Abolition of the English Department" and more substantially in Decolonising the Mind. Yet, the ambivalence of A Grain of Wheat is still fittingly indicated by its current shelf number in Makerere library, where it is classified under the local attribution for African literature in English, under the 827 DDC classification that was originally meant for "English Humor and Satire."

I offer this close-reading, not only to illustrate the presence of the library as an institution of colonial control in the writing of the time I am focusing on in this paper, but also to extend the idea of reading for background to a critical reengagement with library cataloging that has fallen into the background of decolonization debates.

FOREGROUNDING AFRICAN LITERATURE IN THE CATALOG

During Ngúgi's time working at the library of the East African Agriculture and Forestry Organisation he was writing stories, which "were published in various issues of Penpoint and in other regional magazines, like Nilotica. Some, including "The Black Bird," were adapted for radio in 1962, along with the one-act drama The Rebels, and broadcast by the Uganda Radio" (Birth 111). Little magazines, oneact dramas, and radio plays are, indeed, some of the key materials sites of African literary history, yet their placement, or displacement, in the library catalog has meant that they are not easily found by browsing students (even in online catalogs): a case in point is the original Makerere publication of Ngúgi's play, The Black Hermit, mentioned above.

Butitis not only at the level of a student browsing the library stacks that these ontologies fragment and obscure African literature as a field. Little magazines like Penpoint and Nilotica were, as we well know, crucial sites of the development of African literature in the mid-20th century. Already in 1967, at a conference on the Bibliography of Africa, Valerie Bloomfield pointed out: "In Africa where so many periodicals are "little" they provide the main vehicle for the publication of creative writing, criticism and literary news and information. Black Orpheus and Transition are well known, but titles such as Darlite (Dar es Salaam), Okyeame (Accra), Mbari Newsletter (Ibadan), Voix muntu (Congo-Kinshasa), Chemchemi newsletter (Nairobi), seldom appear in the indexes to little magazines [in library catalogs]" (224). Despite recent critical interest in African little magazines, these materials (and their Web 2.0 offspring) continue to cause problems for librarians to this day. Little magazines, like other kinds of ephemeral literatures, such as chapbooks or market literatures, are difficult to acquire and difficult to "enforce deposit" (Bloomfield 224) in national bibliographies. As informal literatures, these materials sit uncomfortably in formal catalogs."

This speaks to a much wider problem that such ephemeral literatures have in entering circulation in the "world republic of letters" (Casanova 2007). If national bibliographies do not capture the metadata of these materials, it is very difficult for them to land in the major literary deposits of world literature, such as the Library of Congress Bibliography or the World Catalog. I have discussed the ease with which the LC subject headings and catalog cards, and even the "standard books" of Western literary education, flowed into use at a library like Makerere's, but the flow of African materials back into the machinations of world literature has not enjoyed the same ease. In 1966, the same year that the category "African Literature" appears in the LC subject headings, the LC launched the National Program for Acquisitions and Cataloguing in, amongst other global locations, Nairobi. The Nairobi branch covered the entire East African region and sought to source materials collected by local libraries, mostly via their accessions lists. Jerry R. James, who worked at the Nairobi branch during its first year of operation, wrote: "Most helpful and current, or relatively so, are the Accessions lists issued by University College, Dar-es-Salaam on a monthly basis; and Makerere University College, Kampala on a quarterly basis" (James 77). However, these accession lists and the Nairobi branch's "acquisition of materials [was] limited to monographic publications" (James 75).

If we recall that Makerere University's library had, in 1966, twenty-two years of library organization based on the DDC and (from around 1956) the Library of Congress subject-headings, and that both systems are heavily biased in their literary ontologies towards the materiality of the book monograph, it is probably not surprising that the librarians there were largely only adding monographic books to their accessions lists. In this we can concretely understand the extent to which an organizational model, developed to describe a particular, American, collection, can have consequences for the global circulation of an important body of literature.

That said, massively important materials for the field of African Literature of the Makerere collection are the numerous MA and PhD dissertations on oral forms of the East African region. It is reassuring that Makerere University Library has embarked on a digitization project, which is creating an online institutional repository for all student work housed in the library-even if older materials (from before the 1990s) are still awaiting digitization. The browsing of this material is now choreographed on the institutional repository page" for the School of Languages, Literature and Communication. This means that the materials collected there cover both oral forms, such as Isaac Tibasiima's The Concept of Power in the Praise Poetry of the Batoro (2010); written forms, such as Aisha Nansubuga's Representation of Trauma in Selected Ugandan Short Stories (2021); and a variety of studies on ephemeral and other forms such as Web 2.0 writing, radio drama, and newspaper literatures. The curation does not, of course, provide a legible corpus of the literary alone, but at least such online collections start to break from the regime of the LC subject-headings and DDC call numbers discussed in this paper.

Interestingly enough, at the time that Makerere Library was introducing the LC subject headings in 1956, University College London's deputy librarian, Kenneth Garside was flying in the face of LC and DDC international standards, by organizing the UCL collection in "subject libraries [to better] meet the needs of teaching departments by reflecting faithfully the current academic approach to the literature of each subject" (170). Had this model been adopted at Makerere in the late 50s, we can imagine that the state of the subject of English literature would have meant the same sort of invisibility for African literature in the stacks as we see via the DDC and LC systems. However, we might speculate that such an organization would have led to quicker change in the library after the decolonization of the literary curriculum, when the English department of East African universities like the University of Nairobi and Makerere became renamed as departments of Literature and opened up explicitly to include local oral traditions." Instead, although the curriculum and the stacks are, as I have argued, so closely linked, the politics of the one did not quickly alter the practices of the other-even if it did so minimally over time. We could say that the digital collections of Makerere, curated as they are under the logic of departments in the university, follows Garside's simple rule.

Yet, on the whole, small-scale and local interventions such as the digitization project at Makere are not enough to address the larger issues of literary ontologies, metadata, and their continued shaping of the field of African literature. In my larger work, of which this article forms a part, I outline the problem of bibliographic metadata and the field of African literature from the early 19th century to the present. This is done in order to develop a Linked Open Database that connects African literary forms that have not been curated for visibility in the major bibliographic standards of the West. By describing the ways that the literary archive has historically obscured the links and connections in African literature, this work hopes to find concrete ways to repeal the impact of the historical catalog on the field."

ACKNOWLEDGEMENTS

This research is supported by the European Research Council (ERC) under the European Union's Horizon Europe research and innovation programme. Grant agreement No: 101097763. I am grateful to Isaac Kiiza Tibasiima (Department of Literature, Makerere University) and Patrick Sekikome (Research Librarian, Makerere University Library) for helping me to access the materials I needed in incredible haste. Thanks also go to Amy Philips, cataloging policy specialist at the Library of Congress.

Footnote

NOTES

1. Karin Barber's detailed engagement with the methodological implications of such "texts" is in The Anthropology of Texts, Persons and Publics: Oral and Written Culture in Africa and Beyond is the most well-known and discipline-changing example.

2. Tanzania gained independence in 1961, Uganda in 1962, and Kenya in 1963.

3. Incidentally, this institution was also affiliated with University College London through a recommendation by the Asquith commission on Higher Education in the Colonies in 1945 (Francis-Brown and Francis 27), four years before the same commission's recommendation was approved for Makerere. University College of the West Indies became an independent university in 1962, alongside national independence for Jamaica and Trinidad and Tobago (Francis-Brown and Francis 27) and just one year before Makerere's independence from UCL.

4. The resemblance is slight, and I have not been able to confirm that this was architecturally the case.

5. This ambition was never fully realized. The library did, however, become a national if not regional repository in 1957 (Larter; Macpherson 93).

6. 131/1 Letter from Archivist, Central Government Offices of the Chief Secretary of Kenya. To Mr Holdsworth, Librarian at Makerere. 10 Aug. 1956. Response on 15 Aug.

7. Indeed, the literary ranges have not changed substantially since the first edition in 1876. For example, 892 Semitic languages (1876) has been expanded to "Afro-Asiatic languages, Semitic literatures" (current); 893 Hamitic and other languages becomes Non-Semitic Afro-Asiatic literatures; and 897 North American Indian literatures becomes "North American native literatures." Dewey 1876, 119; DDC23 800 category in https://www.oclc.org/content/dam/oclc/dewey/ddc23-summaries.pdf.

8. An instruction on this matter reads: "Add to base number 398.2089 [folklore range] the numbers following 305.8 in 305.81-305.89 [ethnic and national groups range], e.g. folktales of North American native peoples 398.208997" (DDC 300, 552).

9. This title is classified under both 398 and 808, without cross-referencing, suggesting that two copies of the text were cataloged, differently, by different catalogers at different times. This also indicates how unclear these categories are for librarians working with African literary materials.

10. See, for example, wa Thiong'o, Lo Liyong and Owuor-Anyumba (1968)

11. MARC (which stands for Machine-Readable Cataloging record) is a format for bibliographic data developed in the 1960s by the Library of Congress and implemented late that decade as the first computer-based format for bibliographic metadata.

12. 1 could not ascertain when the 827.9669 code became assigned to Anglophone African Literature at the library. There is some inconsistency here: the MARC records occasionally record 892 as a standard DDC classification, as "other classification number."

13. By the 17th version of what was called "Volume 1" in 1876, we see the use of the 896 category for African language literature, even though that same year, Melvil Dewey published his A Classification and Subject Index for Cataloguing and Arranging the Books and Pamphlets of a Library, in which 896 is dedicated to "Keltic" language literatures (Dewey 1876a, 21).

14. At a research visit to the library in "fresher" week 2023, I witnessed student groups being introduced to the analog catalog.

15. The Library of Congress also has a call number system, which more directly translates its subject-headings to shelf positions, but since Makerere University Library maintained the Dewey Decimal Classification system for its shelving, I will not discuss the LC call number system in any depth.

16. She writes "The growth of the present author index will not provide any space problem before 1955; the compiling of a subject catalog would reduce seating space on the ground floor; the possibility of providing this catalog is dependent on staff increases." Notes on Library development during the quinquennium 1951-1955. J Larter, Librarian. 26 Sept. 1952. Point 2, page 3 "Catalogue."

17. Library reports, 1946-1997, 141/3 University library reports 58-83, Report of the librarian on the working of the library from 1st April to 31st July 1956. No author. Point 3 "subject index."

18. Тат grateful to Amy Philips, cataloging policy specialist at the Library of Congress, for sending me this information. Phillips also informed me that "African studies first appears in the supplement (Jul-Aug. 1960) to the 6th edition of LCSH. Africa in literature appears in the 1956 supplement to the 6th edition of LCSH" (E-mail to the author).

19. Mowery, a theologian by training, wrote a number of articles concerned with the problems of structural invisibility in library catalogs: apart from African Literature, he also published on African History, Mexican Americans, Women in Literature, and Chicano Literature in subject catalogs.

20. LC Classification-Additions and Changes, January-March (as discussed by Mowery 351).

21. Unlike the concerns I raise above about the "Ethnic and national groups" range in DDC, the LC has recently revised its "Demographic groups" structure to prioritize self-identification and, as such, invites applications from demographic groups to be represented in the subject headings. See all documentation at https://www.loc.gov /aba/publications/FreeLCDGT/freelcdgt.html.

22. Ngügi qualifies this statement by indicating that he is not including settler authors like Elspeth Huxley in this claim.

23. The problem of ephemera in the library catalog is itself a major subject that I will take up in the larger study of which this article forms a part.

24. Peter Poole, as Ngiigi reminds us, what the "first and only white person to hang" (71) for the crime of murder of a black person in Kenya.

25. For a fuller discussion of the problem of literary metadata and African little magazines, see Harris.

26. See http://makir.mak.ac.ug/handle/10570/31/browse?type=dateissued.

27. Makerere's Department of Literature acknowledges this development as the work of Pio Zirimu (https://llc.mak.ac.ug/departments/dlit).

28. Follow the African Literary Metadata Project's progress at http://almedaresearch org.

References

WORKS CITED

Archivist, Central Government Offices of the Chief Secretary of Kenya. Letter to Mr Holdsworth, Librarian at Makarere University Library. 10 August 1956. Makerere University Library, Special Collections, 131/1.

Barber, Karin. The Anthropology of Texts, Persons and Publics: Oral and Written Culture in Africa and Beyond. Cambridge UP, 2007.

Bein, Ann. "Cataloguing of Materials in African Languages." Languages of the World: Cataloging Issues and Problems, edited by Martin D. Joachim, The Haworth Press Inc, 1993, pp. 97-114.

Bloomfield, Valerie. "African Ephemera." The Bibliography of Africa: Proceedings and Papers of the International Conference on African Bibliography, Nairobi, 4-8 December 1967, edited by J. D. Pearson and Ruth Jones. Frank Cass & Co, Ltd, 1970, pp. 223-239.

Casanova, Pascale. The World Republic of Letters. Translated by Malcolm DeBevoise, Harvard UP, 2007.

Dewey, Melvil. A Classification and Subject Index for Cataloguing and Arranging the Books and Pamphlets of a Library. Case, Lockwood Brainard, 1876.

+ Dewey Decimal Classification and Relative Index, vol. 1: Tables, 16th ed., Forest Press, 1958.

. "Dewey Summaries 300." Edited by Joan S. Mitchell et al., 23rd ed, OCLC Online Computer Library Center, 2011,<https://www.oclc.org/content/dam/oclc /webdewey/help/300.pdf.

- - "Dewey Summaries 400." Edited by Joan 5. Mitchell et al, 23rd ed., OCLC Online Computer Library Center, 2011,<https://www.oclc.org/content/dam/oclc /webdewey/help/400.pdf.

- "Dewey Summaries 800." Edited by Joan 5. Mitchell et al., 23rd ed., OCLC Online Computer Library Center, 2011,<https://www.oclc.org/content/dam/oclc /webdewey/help/800.pdf.

+ "DDC 23 Summaries" Edited by Joan 5. Mitchell et al., 23rd ed., OCLC Online Computer Library Center, 2011,<https://www.oclc.org/content/dam/oclc/dewey /ddc23-summaries.pdf.

Dipio, Dominica. "Literature's Raison D'être: Disciplinary Limits of the Art-Past and Present." Historicising the Humanities at Makerere: Trends, Patterns and Prospects, edited by Josephine Ahikire, Levis Mugumya, Edgar F. Nabutanyi, and Peter Atekyereza. Fountain Publishers, 2022, pp. 73-105.

Edlund, Paul. "A Monster and a Miracle: The Cataloging Distribution Service of the Library of Congress, 1901-1976." The Quarterly Journal of the Library of Congress, vol. 33, no. 4, October 1976, pp. 383-421.

Foucault, Michel. The Archeology of Knowledge and The Discourse on Language. Translated by A. M. Sheridan Smith, Pantheon Books, 1972.

Francis-Brown, Suzanne, and Peter Francis. "Norman & Dawbarn, the UCWI, and Tropical Modernist Architecture in Jamaica." Caribbean Quarterly, vol. 65, no.1, 2019, pp. 27-56. DOI: 10.1080/00086495.2019.1565219

Furner, Jonathan. "Dewey Deracialized: A Critical Race-Theoretical Perspective." Knowledge Organisation, vol. 34, no. 3, 2007, p. 144-68. DOI 10.5771/0943-7444 -2007-3-144.

Garside, Kenneth. "The Basic Principles of the New Library Classification at University College, London." The Journal of Documentation, vol. 10, no. 4, December 1954, рр. 169-92.

Green, Louise. "Reading for Background: Suyi Davies Okungbowa's David Mogo, God Hunter and 'The End of the World as We Know It." ALT 38 Environmental Transformation: African Literature Today, Boydell & Brewer, 2020, pp. 24-36. DOT: 10.2307/j.ctv105bbzf.6

Green, Marilyn A. and Susan Rathbun-Grubb. "Notes on Operations Classifying African Literary Authors." Library Resources and Technical Services, vol. 60, no. 4, October 2016, pp. 270-77. DOI: 0.5860/1rts.60n4

Harris, Ashleigh. "The Literary Metadata of African Little and Popular Magazines." ALMEDA Pamphlets, 1 Jan. 2024,<https://almedaresearch.org/almeda-method -pamphlets/.

Holdsworth, Librarian at Makarere University Library. Letter to Archivist, Central Government Offices of the Chief Secretary of Kenya. 15 Aug. 1956. Makerere University Library, Special Collections, 131/1.

James, Jerry R. "The Library of Congress Program in Eastern Africa." The Bibliography of Africa: Proceedings and Papers of the International Conference on African Bibliography, Nairobi, 4-8 December 1967, edited by J. D. Pearson and Ruth Jones, Frank Cass & Co, Ltd, 1970, pp. 75-82.

Kisiedu, Christiana O. "Cataloguing and Classifying Non-Western Materials in Ghanaian Libraries." Cataloguing and Classification of Non-Western Materials: Concerns, Issues, and Practices, edited by M. M. Aman, Oryx Press, 1980, pp. 31-73.

Kotei, S. I. A. "Some Problems in Africana Library Classification." The Bibliography of Africa: Proceedings and Papers of the International Conference on African Bibliography, Nairobi, 4-8 December 1967, edited by J. D. Pearson and Ruth Jones, Frank Cass & Co, Ltd, 1970, pp. 138-54.

Kua, Eunice. "Non-Western Language Literatures in the Dewey Decimal Classification Scheme." Libri, vol. 54, 2004, pp. 256-65.

Kyle, Barbara. "Towards a Classification of Social Science Literature." American Documentation, vol. 9, no. 3, July 1958, pp. 158-83.

Larter, J. "Notes on Library Development during the Quinquennium 1951-1955." Makerere University Library, Special Collections, 141/3, 26 Sept. 1952.

Library of Congress. "A." Library of Congress Subject Headings, 44th ed., 2022, https:// www.loc.gov/aba/publications/FreeLCSH/A.pdf.

. "E" Library of Congress Subject Headings, 44th ed., 2022, https://www.loc.gov /aba/publications/FreeLCSH/E.pdf.

___ "Е" Library of Congress Subject Headings, 44th ed., 2022, https://www.loc.gov /aba/publications/FreeLCSH/F.pdf

. "К" Library of Congress Subject Headings, 44th ed., 2022, https://www.loc.gov /aba/publications/FreeLCSH/K.pdf.

- "1." Library of Congress Subject Headings, 44th ed., 2022, https://www.loc.gov /aba/publications/FreeLCSH/L.pdf.

- "O." Library of Congress Subject Headings, 44th ed., 2022, https://www.loc.gov /aba/publications/FreeLCSH/O.pdf.

Library of Congress Subject Catalog Division. LC Classification -Additions and Changes, January-March 1973.

. Subject Headings Used in the Dictionary Catalogs of the Library of Congress, edited by Mary Wilson MacNair, 4th ed., Library of Congress, 1943.

+ Subject Headings Used in the Dictionary Catalogs of the Library of Congress, edited by Nella Jane Martin, 5th ed., Library of Congress, 1948.

. Subject Headings Used in the Dictionary Catalogs of the Library of Congress, edited by Marguerite У. Quattlebaum, 7th ed., Library of Congress, 1966.

Littletree, Sandy, and Cindy Aden. "Centering Washington Tribal Libraries: Establishing the Foundations". ALKI: The Washington Library Association Journal, vol. 38, no. 3. December 2022, pp. 5-8.

Mann, Thomas. "The Research Library as Place: On the Essential Importance of Collections of Books Shelved in Subject Classified Arrangements." The Library as Place: History, Community, and Culture, edited by John E. Buschman and Gloria J. Leckie, Libraries Unlimited, 2007, pp. 191-206.

Macpherson, Margaret. They Built for the Future: A Chronicle of Makerere University College, 1922-1962. Cambridge UP, 1964.

Makerere University Library. Manual of Cataloging Procedures, Makerere University Library, 1972.

Mitchell, Joan S. "DDC 22: An introduction." International Cataloguing and Bibliographic Control, vol. 33, no. 2, April 2004, pp. 27-31.

Mowery, Robert L. "The Classification of African Literature by the Library of Congress." Library Resources and Technical Services, vol. 17, no. 3, Summer 1973, pp. 340-52.

Mutula, Stephen M., and Mashingaidze Tsvakai. "Historical Perspectives of Cataloguing and Classification in Africa." Cataloging & Classification Quarterly, vol. 35, nos. 1-2, 2002, pp. 61-77. DOT: 10.1300/J104v35n01_05.

Nabutanyi, Edgar F., and Isaac Tibaisiima. "A Peripheral Archive or a Discursive Platform of Post-Colonial Uganda? Historicising Trauma in Penpoint." Historicising the Humanities at Makerere: Trends, Patterns and Prospects, edited by Josephine Ahikire, Levis Mugumya, Edgar F. Nabutanyi, and Peter Atekyereza, Fountain Publishers, 2022, pp. 178-200.

Ngügi wa Thiong'o. Birth of a Dreamweaver: A Writer's Awakening. New Press, 2016.

. Decolonising the Mind: The Politics of Language in African Literature. James Curry, 1986.

. A Grain of Wheat. 1967. Heinemann African Writers Series, 1986.

Ngúgi wa Thiong'o, Taban Lo Liyong, and Henry Owuor-Anyumba. "On the Abolition of the English Department." The Norton Anthology of Theory and Criticism, edited by Vincent B Leitch, 2nd ed. 1968. W. W. Norton & Company, 2010, pp. 1995-2000.

Philips, Amy. E-mail to the author, 1 Nov.' 2023.

Schweitzer, Ivy, and Henry, Gordon Jr., eds. Afterlives of Indigenous Archives. Dartmouth College Press, 2019.

Urban, Greg. "Entextualisation, Replication, and Power." Natural Histories of Discourse, edited by Michael Silverstein and Greg Urban, Chicago UP, 1996, pp. 21-44.

Wisecup, Kelly. Assembled for Use: Indigenous Compilation and The Archives of Early Native American Literatures. Yale UP, 2021.

Witherall, Julian W. "Bibliographic Programs on Africa in Washington, DC, at the Library of Congress and the African Bibliographic Centre." The Bibliography of Africa: Proceedings and Papers of the International Conference on African Bibliography, Nairobi, 4-8 December 1967, edited by J. D. Pearson and Ruth Jones, Frank Cass & Co, Ltd, 1970, pp. 126-34.