Introduction

Knowledge organization systems (KOS’s) have been criticized for being products of their particular moments in time, most notably in the vocabulary used for subject representation, especially regarding groups of people. It has also been established that the hierarchical, mutually exclusive structure used for most KOS’s can subordinate, scatter, erase or ghettoize topics, resulting in access problems (^{Olson & Schlegl, 2001}). The bibliographic KOS’s that have been the objects of the most derision have been the Dewey Decimal Classification (DDC) and Library of Congress Subject Headings (LCSH). Dewey (¹⁹⁸⁷, p. 4) openly recognized the shortcomings of his system, writing,

many minor subjects have been put under general heads to which they do not strictly belong…The rule has been to assign these subjects to the most nearly allied heads, or where it was thought they would be most useful. The only alternative was to omit them altogether.

but for most of its history, the Library of Congress remained silent on the reasoning and philosophy behind creation and application of the list of subject headings despite much criticism.

In the most famous critique, Sanford ^{Berman stated in 1969} that the Library of Congress Subject Headings represent a white racist imperialist point of view, which can prove embarrassing in an African university library (^{Berman, 1969}, p. 695). Later, in 1971, Berman wrote his carefully studied and venomous Prejudices and Antipathies, where he identified multiple terms and syndetic relationships where LCSH was offensive, prejudicial, or exclusive. He wrote that

the LC list can only “satisfy” parochial, jingoistic Europeans and North Americans, white-hued, at least nominally Christian (and preferably Protestant) in faith, comfortably situated in the middle- and higher-income brackets, largely domiciled in suburbia, fundamentally loyal to the Established Order, and heavily imbued with the transcendent, incomparable glory of Western civilization» (^{Berman, 1971}, p. 3).

This creates what ^{Fox and Reece (2012}) call «conceptual violence». which they consider harm to users «through linguistic or structural misrepresentation» (p. 377).

In response to vocal public comment on shortcomings and consequences of biased representation and misleading structure, LC began allowing outside feedback in decisions. Until 1984, users of LCSH were not even privy to the rules and instructions for application. Once the Subject Cataloging Manual: Subject Headings was published, catalogers were able to see justification for the assignment of terms, which allowed them to tailor records to users rather than blindly guess. In the past, LC held fast to the power it held over the application of subject headings but realized that the appliers and users of LCSH could provide much-needed context for adjustments to headings.

An alternative interpretation of Berman’s statements might be that not every library knowledge organization system/controlled vocabulary must necessarily be considered inadequate for the mainstream of users within the society that created the system, nor embarrassing from a white racist imperialist view, since the system is based on a reflection of the works (literary warrant) and the intention (the classificationist’s goals) of that mainstream community. However, it should be recognized that in a given society it is hard to find a homogeneous group of users with no dissidents. ^{Fox and Reece (2013}) describe how standards governing bibliographic control create violence at the borders of categories, as they create constraints that remove any possibility for inclusiveness. Any given system that is deemed adequate for a given society may be inadequate for those dissidents and other groups of users assigned either to mainstream groups in different societies or minority groups in the given society. When the knowledge of subordinate cultures is made inaccessible, it can lead to epistemicide or the «killing, silencing, annihilation, or devaluing of a knowledge system» (^{Patin et al., 2021}).

Discrimination against and misrepresentation of marginalized groups in knowledge organization systems did not seem to be a desirable topic to research until recently. As yet not enough has been written on the problems of one-dimensional marginalized groups nor is it at all clear what kind of problems could affect multi-dimensional marginalized groups, also known as intersectional groups. While several academic works have studied individual one-dimensional marginalized groups in library knowledge organization tools, the study of intersectionality in library knowledge organization systems is relatively recent (see the pioneer studies by de la ^{tierra, 2008}; ^{Hogan, 2010}; ^{Olson, 2002}; ^{Martínez-Ávila et al., 2012}; ^{Fox, 2016}; ^{Fox et al., 2017}; ^{Moura, 2018}a). The minoritization vs. universalization debate has been a complex issue in the representation of unprivileged groups in knowledge organization systems. In our study, we review this issue and other concepts such as essentialism and the revision of knowledge organization systems in the light of intersectionality. Methodologically, we can consider it as a review article or synthesis article. We give special importance in our analysis to the Latina example by knowledge organization intersectionality pioneer tatiana de la tierra and the recent Latin American literature in Library and Information Science indexed by the Brazilian bibliographic database BRAPCI.

Intersectionality

Intersectionality, also called «double jeopardy» or «multiple jeopardy» (^{King, 1997}), names the intermeshed oppression systems experienced by those who belong to multiple identity categories that cannot be pulled apart into individual variables and separately addressed. Because it relates to the categorization of human groups, it is of great concern to knowledge organization. Race and gender-black women specifically-made up the original «intersection» of the concept, but its theoretical value has extended to other oppressions such as class, sexual orientation, age, disability, citizenship status, and many more. Whiteness is not exempt from intersectionality, as it «has always been fractured by class, gender, sex, ethnicity, age, and able-bodiedness» (^{Alcoff, 1998}, para. 15). Although these additional oppressions enrich a sense of inclusivity of marginalized populations, they still end up perpetuating some problems endemic to the categorization of groups of people: essentialism, the shifting boundaries of social groups, the definition groups as a whole, and identity versus biology. Additionally, as the number of identity categories increase, the groups become smaller and more specific, which brings up methodological problems of how to cater to each specific population.

Intersectionality has been used in the Library and Information Science literature as a theoretical framework or methodology for the study of several topics. In the Brazilian case, ^{Pinheiro and Inomata (2022}) identified the intersectionality of gender and race among the key aspects studied in master’s theses and doctoral dissertations on feminism in LIS. Also related to these types of works, ^{Gaudêncio et al. (2018}) conducted an intersectional analysis to the Brazilian Biblioteca Digital de Teses e Dissertações database (BDTD), focusing on markers of social difference in the representation of knowledge in action. Maria Aparecida ^{Moura (2018}b) considered intersectionality, and other concepts such as gender, power, and taxonomic reparation, in her analysis of two European knowledge organization systems. This was a very good application of intersectionality to the specificities of knowledge organization. In a more recent study, ^{Moura (2024)} connects intersectionality to other concepts such as coloniality and social differentiation in Brazil. Social phenomena such as the coloniality of being, knowledge, power, and seeing, capitalism, racial hierarchy, and others are discussed as some of the roots of the violence in knowledge organization systems. As a response, Moura presents an Intersectional Thesaurus: he COEXISTENCE - Thesaurus of Intersectionality and Decolonial Issues: Black Studies, Gender, Sexuality, and Feminist Studies. ^{Viana and Carrera (2019}) used intersectionality in the analysis of the visibility of black female youtubers. ^{Carrera (2021)} discussed the application of intersectional roulette method for communication studies. ^{Oliveira el at. (2021}) used intersectionality in the analysis of documents and interviews related to people living by dam failures. ^{Felix and Paulla (2021}) also used intersectionality, focusing on racism, to analyze hate speech targeting Brazilian black politician Talíria Petrone. ^{Silva (2021}) discussed intersectionality in Library and Information Science mainly from a decolonial perspective.

^{Honorato and Honorato (2021}), in a very interesting work, discussed several issues related to intersectionality, representation, and essentialism, comparing the concepts of intersectionality, crossroads («encruzilhada» in Portuguese, see ^{Martins, 1997}), and exuzilhada (^{Silva, 2018}, a non-linear concept that refers to Exu, orisha messenger between temporalities, places, and worlds). ^{Araújo and Machado (2022}) applied intersectionality to de development of a collection considering authors with disabilities and other inclusions. ^{Hayashi et al. (2022}) considered intersectionality in the analyses of podcasts in the field of Library and Information Science that deal with Diversity, Equity, Inclusion and Accessibility (DEIA). ^{Romeiro and Silveira (2022}) considered intersectionality in the analysis of the publications in one of the main academic meetings in Brazil (ENANCIB). ^{Romeiro and Silveira (2003)} also presented an analysis of two Brazilian thesauri on gender issues drawing on intersectionality and decolonial theories (another good application in knowledge organization). ^{Pires and de Paula (2023}) studied the intersectionality of gender-race-class oppressions in the library profession, interviewing alumni of the library science undergraduate program of Universidade Federal de Minas Gerais, Brazil. ^{Sampaio et al. (2023)} also studied the librarians’ professional profile through the lens of intersectionality, focusing on gender and age. ^{Côrtes and Silva (2023}) discussed intersectionality, focusing on black feminism, in the context of mediation of information. ^{Silva and Sales (2024a)} studied intersectionality in Library Science, focusing on issues such as gender, race, and class, and drawing on the contributions by Audre Lorde and other key authors to the topic. ^{Silva and Sales (2024b)} continue the discussions, this time mainly drawing on the contributions by Patricia Hill Collins, and discuss the aforementioned concept of crossroads (as reported by Honorato and Honorato, 2021). The authors advocate for an intersectional librarianship for the resolution of social problems. ^{Silva et al. (2024)} considered intersectionality in the study of microaggressions related to race, gender, sexuality, social class, disability, and religion in LIS education. ^{Hayashi and Rigolin (2024a)} discussed intersectionality, focusing on gender, race, and parenting, and applied it to a bibliometric study of the scientific production of black women during the COVID-19 pandemic. Elsewhere, Hayashi and Rigolin (^2024b) also discussed intersectionality in the development of DEIA indicators.

Other recent works in Latin America such as those by ^{Duque Cardona (2022}), Duque-Cardona and Restrepo-Fernández (²⁰²²), ^{Schifrin (2022}), and ^{Lessa and Crosara (2024}) have also acknowledged intersectionality. ^{Cardoso et al. (2023}, p.14) even close their work stressing that «it is essential to consider intersectionality, understanding that oppressions are intertwined with other social markers, such as race, social class, sexual orientation, among others».

In knowledge organization, intersectionality can reinforce or exacerbate tension about categorization. ^{McCall (2005}) finds a spectrum of attitudes toward categorization in feminist theory that underpins intersectionality frameworks. At one end, categories are rejected (anticategorical complexity), and at the other end they are used strategically for political ends (intercategorical complexity). A third approach, called intracategorical complexity, epistemologically falls in the middle and means that categories are used in effect as placeholders for describing intersectional locations. Many feminist theorists, notably ^{Spelman (1988}) and ^{Frye (1983}), have taken issue with categorization, believing that categorization «leads to demarcation, and demarcation to exclusion, and exclusion to inequality» (^{McCall, 2005}, p. 1777). Some scholars believe eliminating categories could lead to equality, while acknowledging that categories are politically necessary and linguistically unavoidable. Spelman has specifically questioned whether members of biological or social groups truly have very much in common, or as ^{McCall (2005}, p. 1777) argues, that «language…creates categorical reality rather than the other way around». ^{Butler (1990}) and other postmodern, «postfeminist» scholars have also called into question the stability of «woman» as a category, undermining the use women as an object of study because they are «too irreducibly complex to categorize» (^{McCall, 2005}, p. 1773). On the other hand, without categories, the experiences of the unprivileged groups are erased, which leads to the adoption of provisional or strategic labels. A type of strategic essentialism, popularized by Gayatri Spivak, advocated politically grouping together to gain power, though she later rejected it (^{Olson & Fox, 2010}). As a midway, the concepts of crossroads and exucilhada have also been proposed as non-essentialist alternatives for the representation of experiences and realities through operators such as analogy and dislocation (^{Honorato & Honorato, 2021}).

The problem of intersectional categories for classification

In KO, the notion of oppression can also be related to principles of classification, categorization and linguistic representation. A 2001 meta-analysis analysis of KO literature written since the 1970’s identified 31 identity-related topics found to be affected by four different types of bias in knowledge organization standards. First, the topic can be treated as an exception, as if its existence strays from the norm (i.e. «women as physicians»). Also, the topic can be «ghettoized», or exiled to a class which removes it from a broader swath of categories. Third, the topic can simply be omitted, as if it does not exist. Fourth, it can be lost in a structure that does not follow the rules of the classification scheme (i.e. nonexistent «see also» references to related topics), and finally, the terminology can be biased (i.e. «lunatics») (^{Olson & Schlegl, 2001}, p. 65), which can be a form of intellectual violence. To add to those, classifications can also commit cultural imperialism by subordinating or delegitimizing groups, making moral judgments, reinforcing stereotypes or dehumanizing groups (i.e. the classification of women and gypsies into the heading of «Customs Costumes Folklore» in DDC). Another way of erasing difference is to lump library patrons into one essentialized group under the name of «users», «patrons», or «the public» (^{Olson, 1996}). If a searcher cannot find information, finds information in a judgmental hierarchical, or finds it represented in a demeaning fashion, it may be perceived as truth. The consequences of this can be great. While inanimate and categorically ambiguous objects like rugs or sea sponges may not care where or how they are classified, groups of people do care.

Another complication arises where race and race intersect or gender and gender intersect. ^{Furner (2007}) describes the then most recent update to DDC’s Table five, now named «Ethnic and National Groups» which first appeared in the eighteenth edition. The previous edition lists three «basic races» and then lists possible mixtures of those basic races. As problematic as that may be, the 22nd edition eliminated race altogether in favor of ethnic group, causing Furner to comment, «the human population that are typically referred to as “races” are no longer available as subject matter for writers» (p. 156).

A final problematic area with intersectional identities is the notion of facet order. Facets may allow more topic coverage, but in the strict linearity of library shelves and the unforgiving nature of hierarchy, one identity category must come first, which then dictates how topics may be scattered. Hope Olson introduced the problem of intersectionality in library classifications as a matter of differentiation in hierarchy, what Aristotle called differentiae, and equality (^{Olson, 2002}, p. 173):

The way gender operates in conjunction with other discourses of power is hidden by efforts at equal treatment…By regarding each as monolithic and as parallel, differences within as well as between the two are erased. Another result of equal treatment is the privileging of some differences over others. Library classifications are linear in that they line books up in order on shelves so that library users can browse. As a result, it is not possible to gather all aspects or facets of a work simultaneously. Works are gathered by one facet and then subdivided by another and so on, creating a hierarchy. As a result, one facet is the primary point of gathering and others are not gathered in one place.

This aspect of classification was also pointed out by Elizabeth Spelman, who noticed that the picture of what differentiates one person/group from another or what the two have in common will vary according to how the classes are ordered (^{Spelman, 1988}, p. 144). If gender is first and ethnic background comes next, then all women and men will be together, but not all Hispanic-Americans, Afro-Americans, Asian-Americans, etc. which will be in at least two different places. In another example, Olson explains this problem in relation to the universalizing/scattering issue: «Works on lesbians that specify other characteristics such as race or ethnicity will be even further dispersed - if they find it all» (^{Olson & Ward, 1997}, p. 27). This means that classificationists must choose which facet, what aspect of the personality of a given group, will remain hidden or even ignored throughout the system.

Struggle of marginalized groups in library catalogs

Both identity groups in an instance of intersectionality could individually be considered marginalized groups and marginalized topics (or dissidents) within the dominant society and its library knowledge organization systems. Those groups have been marginalized, pushed away from the situational reference point and forced to be «other» by mainstream culture (reflected through books by literary warrant) and standardized control vocabularies (such as LCSH and DDC). As ^{Olson (2000}, pp. 57-60) explained,

the tradition of literary warrant at the Library of Congress has been the same for both classification and subject headings. It has echoed the mainstream and rejected the margins… Marginalization of a topic is the process of placing it outside of the cultural mainstream-making it “other”.

A further step from the center of the system would be the intersection, obviously more marginalized than its individual components and with different characteristics and problems too.

One of the clearest common problems in knowledge organization systems, as noted by ^{Olson and Schegl (2001}), is biased terminology since systems are almost always written using the language of the mainstream. What is pejorative for a given community is not necessarily so for another. But with certain groups of users, qualifying terms seem to be deliberately chosen to serve as an alert to society and to ensure their isolation. When these problems occur, the social consequences for users can be psychologically devastating. One example of biased terminology is underscored by the case of a library card with the heading «LIBRARIES AND THE SOCIALLY HANDICAPPED» (changed from «LIBRARY SERVICE TO THE CULTURALLY HANDICAPPED») to cover the topic of access to information by a Latina/o in 1972, reported by Marielena ^{Fina (1993}, p. 269). According to the library card, being Latina meant be socially or culturally handicapped, and thus an impressionable user might be made to feel. However, the worst part might be the subjects she was lumped in with, the whole set of subjects and books included in the socially handicapped class that shared her identity. Again, what could be offensive for the mainstream might also be offensive for different marginalized groups. There is not one sole division between «mainstream group» and «others group,» but as many divisions of as many groups as others exist. This problem is even more delicate if intersectionality is considered.

Concerning the gathering of different others, this is also a problem of an inappropriate structure of the standard. Another example of an, at minimum, inappropriate structure with negative consequences might be, as Richard Smiraglia points out, «when a gay adolescent searches for literature to help understand and finds that it all falls under “perversion” then we have oppressed yet another youth» (^{Smiraglia, 2006}, p. 186). Another example is given by tatiana de la tierra when talking of a seventeen-year-old girl being shocked after asking for material about homosexuality at New York Public Library and finding notes such as «see Deviancy» or «see Pathology» (^{de la tierra, 2008}, p. 94). More examples on this topic can be found in the literature, as ^{Olson stated in her work The Power to Name (Olson, 2002}, p. 2). Julia Penelope, in the process of her coming out, looked for information at the Miami Public Library in Biscayne Park. She writes: «when I discovered the books on “homosexuality”…I read passages that described “mannish” wimmin with short, cropped hair…And I thought to myself, “That’s me!”…I knew what I was» (^{Penelope, 1989}, p. 59); Judy Grahn at twenty-one looked for information on homosexuality and lesbians at a library in Washington D.C. to learn who she might be; she found that the books on such a subject were locked away: «Only professors, doctors, psychiatrists and lawyers for the criminally insane could see them, check them out, hold them in their hands» (^{Grahn, 1984}, p. xi); and more. While the first case is related to a biased assimilation of alien ideas from a book, the second case is a problem related to invisibility of the books. The first example might have happened even if the book handled the subject properly but it was lumped together with other books treating the topic as a disorder. The second example might have happened even if the books were not locked away but classified under some other unrelated topic as a result of the subject having no name. In every case there would be miscommunication between books and users, and an act of aggression perpetrated by the knowledge organization system/library service toward the marginalized users.

Other examples on this topic include, for example, ^{Wolf (1972}) documents how both the Dewey Decimal Classification and Library of Congress classifications lumped «gayness» with crime and sexual disorders, prostitution and pornography, disorders of character, rapists, seducers, and perversions through context and references. This was in contrast to the nuclear family and sex within marriage, accepted and set up as the norm. In response to this, C. Sumner Spalding, then Assistant Director for Cataloging of the Library of Congress, replied to Wolf that the Library of Congress does not establish usage, but reflects it (^{Wolf, 1972}).

This topic is an especially sensitive one, and library responsibility is great because, as Greenblatt (1999, p. 87) pointed out:

In a society in which public acknowledgment of one’s nonconforming sexual and emotional orientation is uncomfortable at best and dangerous at worst, many turn to books rather than people for information concerning their surfacing identities. As such, the library has often been one of the first choices as a source of information on this subject.

However, as ^{Olson (2000}) suggests, librarians’ responsibility and room for their improvements can sometimes be shattered by the concept of literary warrant that rules the development of the standard.

Although the categories of problems identified by Olson and Schlegl can occur in relation to a variety of topics and to a variety of groups of users, some categories of problems seem more likely to happen with some groups than with others due to internal disagreements. For instance, some LGBTIQ+ will inevitably have to struggle more often with problems of ghettoization of the topic and inappropriate structure of the standard, since the communities’ lack of consensus is more evident and thereby the domain goals and values are less clear. In this vein, some authors such as Grant ^{Campbell (2000)} and ^{Christensen (2008}), have studied the pros and cons of the minoritizing and universalizing stances in the classification process for the queer community. Christensen’s work on minoritization vs. universalization will be revised below due to its importance for the topic.

In his introduction, Christensen recalled Grant Campbell’s quotation on the importance of bibliographic access tools for the representation of certain groups: «survival within a marginalized group depends on the regular and frequent subversion of traditional classification categories» (^{Campbell, 2000}, p. 127). Assuming that the Library of Congress Subject Headings (LCSH) and the Library of Congress Classification (LCC) include some of the most traditional categories used in libraries, his work studied the subversion of these schemes for lesbians and gay men as examples of traditionally marginalized groups. Concerning the LCSH, this topic was thoroughly revised by Ellen ^{Greenblatt in 1990}, although nobody followed up for almost two decades. Similar work has not been done regarding the LCC, perhaps, because its underlying structure and terminology are not quite so visible to library users, their perception is given underrated importance. In 2008, Christensen studied the evolution of Greenblatt’s suggestions in both schemes since then. In 2011, Greenblatt revised her own work too. However, the underlying philosophy behind most of Greenblatt’s proposed changes was the minoritizing view, one of the stances in the homo/heterosexual representation dilemma of minoritizing vs. universalizing. Greenblatt (in 2011, p.213) justified this stance as follows:

As this essay examines the treatment of concepts pertaining to groups under the LGBTIQ umbrella, it will of necessity assume a minoritizing approach to the topic taking the stance that marginalized identities should not be invisible and that controlled vocabularies, such as LCSH, should enable users to easily find the information they seek using culturally sensitive and relevant terminology.

According to ^{Campbell (2000}, p. 129), the interpretation of this binary in library classifications is that

the universalizing tendency will tend to treat explicit subject headings with suspicion…The universalizing approach implies that the explicit presence of a topic in a subject access system implies a deviation from the norm. The minoritizing view, on the other hand, may well argue…for “visibility at any cost;” “I’d rather have negative than nothing”.

It seems that there is no easy solution, and while some authors like Greenblatt advocate for the minoritizing view, other authors like ^{Campbell (2000)} seem to prefer the universalizing view. In practice, this is a matter of which facets to highlight, and, according to Christensen, the final decision of on this should be taken according to how the community categorizes itself and not only on current usage (literary warrant). On the other hand, many of the problems remain unresolved in the standards. Most of the problems that Greenblatt detected in the LCSH and the LCC had not been corrected for more than 20 years. Neither the minoritizing nor the universalizing view is being considered in the scheme. Moreover, when a bias is corrected, it is assumed that the Library of Congress did not take the suggestions of any collective or expert into consideration but mainly relied on literary warrant (that, as pointed out by ^{Martínez-Ávila and Budd, 2017}, follows an empiricist approach). This can be interpreted as reinforcement of the authoritative discourse of the Library of Congress and the social reflection of the library market as the true authority when developing standards, in opposition to the theoretical and political arguments that are considered in the pragmatist domain-analytic approach (^{Hjørland, 2017}) in which the homosexual community is not considered a homogeneous group.

The problems of the Library of Congress (and other organizations in charge of knowledge organization systems) when «correcting» universal standards can be understood at with regards to the structural aspect. It might be argued that the Library of Congress cannot satisfy everybody when developing the hierarchical structure of the standard as universal standards should follow universalizing views. However, from a post-structural point of view, the successful search for universality is impossible and therefore it is impossible to eliminate every kind of bias and adopt a universalizing view. The solution would be a reversal of the binary, a minoritizing view in an alternative standard where the norm is the local marginalized group that will use the scheme, a subversive alternative where the inevitable bias of the marked representation of the one would only be valid for this context and could be linked to other standards that would be valid for different contexts through switching languages since its underlying structure and terminology is not quite so visible to library users. From an intersectional point of view, any dominant position will always be adopted. On the other hand, linguistic problems should be easier to resolve since they do not affect structure, and in the case of ^{Greenblatt’s analyses (1990}, ²⁰¹¹), have been ignored for a long time, as have other marginalized groups represented in the scheme. This problem exposes a basic failing in the Library of Congress revision process based on literary warrant and not directly on context. The library access tool functions passively instead of being proactive, while at the same time its influence on society has been underrated from its conception. In other words, the LCSH and the LCC as systems suffer from a passivity that ignores the cultural and social responsibility of Librarianship and Information Organization tools and practitioners.

Finally, in the context of the minoritization vs. universalization debate, the omission of a topic might also be considered an extreme case of universalizing. In the same way that there is no topic on people who breathe, the existence of a topic that cannot deviate from any norm while it has no representation at all would not be noticed. Visibility at any cost, either by marked representation or explicit naming, might also be a good idea for reversing representation and shifting bias. One example of this phenomenon is the community’s proud adoption of external terms intended to cause injury to the group. However, although the subversive use of some terms may have some effect on society and on user perception, the use of such subversive terms, even when widely used by the community in the literature, is rarely reflected in library tools and catalogs even if they are based on literary warrant, as shown by tatiana de la ^{tierra’s pioneer study on intersectionality (2008}). As tatiana de la tierra noticed in her work «Latina Lesbian Subject Headings: the Power of Naming», «HOMOSEXUAL pointed me to books about dysfunctional identities resulting from overbearing mothers. HISPANIC in association with LESBIAN led nowhere» (^{de la tierra, 2008}, p. 95).

As a lesbian, Latina, user, author, and librarian, tatiana de la tierra observed some of the consequences and responsibilities of these problems firsthand. Unlike the works of Olson and Christensen, which do not focus so much on any particular use or local catalog but more on the standards themselves (the classificationist’s responsibility), de la tierra’s work is more concerned with the application of the standard and the way end-users perceive the described books in a particular classified library or collection (the classifier’s responsibility). As a librarian and a writer, she analyzes the terms used by the communities she is involved with, the terms found in the specialized literature on the topics (giving examples) that she might use too, and contrasts the inappropriate official alternatives given by the literary warrant based Library of Congress (working from the 27th and 28th editions of LCSH) to the local application of the University at Buffalo’s (UB) BISON catalog. The organization and classification of LGBT materials have been described as challenging by authors such as Liana Zhao, head of the library at the Kinsey Institute, in part because sex generally had been a taboo topic for much of social history, queer sex even more so (^{Zhao, 2003}). To this Patrick Keilty added that: «indeed, it is precisely because queer belongs to the categories of perversion that it transgresses the traditional boundaries of desire» (^{Keilty, 2009}, p. 242). However, de la tierra does not consider the minoritizing vs. universalizing issue on LGBT introduced by ^{Campbell (2000}, p. 128) and further developed by Christensen, as she implicitly adopts a minoritizing, almost essentialist, view for most of her statements is needed to signal the additional state of oppression. Examples of this minoritizing instance could include the distinction between different denominations of latin@s and the highlighting of the importance of having a name, a marked representation:

Because even worse than inappropriate terms is the lack of subject headings. In these cases, if titles don’t contain magical keywords, the books are effectively erased from catalogs. To not name is to eradicate, to make invisible. It is like banning a book that no one ever knew existed to begin with» (de la ^{tierra, 2008}, p. 100).

De la tierra here lays out the problem of topic omission as something that is even worse than ghettoization or an inappropriate structure, and therefore, those problems typically experienced in intersectionality cases have worse consequences than do those experienced by one-faceted marginalized groups. In this view, all facets must be represented in the system.

The importance of naming in de la tierra’s work is clear. One of the first aspects that gains attention on tatiana de la tierra’s work is the lower case spelling of her name. This spelling might be interpreted as a social device for demanding voice or visibility in her condition. However, as Olson pointed out while studying the similar case of African American woman bell hooks, who also spells her name using lower case letters, this device is systematically muted in the subject heading by the equalitarian search for neutrality and order of the standard. Specifically, the application of the «Anglo-American Cataloguing Rules 2nd edition, 1988 revision (AACR2R) Appendix A. Rule A.2.A1,» indicates that the first word of each heading and subheading must be capitalized. Consequently, tatiana de la tierra’s right to name herself as she wants is rejected by the standard that defines her, something that might be considered fundamental for her and her discourse, as David ^{Abalos (1999}, p. 18) notes:

The language and discourse of liberation and transformation recognizes and honors the right of all people and groups to name themselves. This is not about a fad, being politically correct, or a commercialization or difference; to name oneself and one’s community is at one and the same time a personal, political, historical, and sacred act.

Names and the right that communities should have to name themselves is something that is constantly repeated through de la tierra's article:

To name, to categorize and classify, to label and brand, to make a linguistic determination, to signal, to define, to say, “this is the word, these are the words that will represent you”-this is a powerful thing. Those of us who had to learn another language and culture know about the power of naming, of being named, and of making words our very own (de la ^{tierra, 2008}, p. 95).

The relationship of respect and reflection of this aspect to faceted classifications was pointed out by Campbell: «If faceted organization schemes are to work, however, we need to know more about gay and lesbian users, and how they categorize themselves and their information sources» (^{Campbell, 2004}, p. 109).

Related to this, the problem of subjects with no names (i.e. not named by any authority) was also considered by Cutter, who even used similar examples related to ethnicities and nationalities:

Our literature can not be treated satisfactorily. It is never called United States literature, and no one would expect to find it under United States. On the other hand the name American properly should include Canadian literature and all the Spanish literature of South America. It is, however, the best name we have» (^{Cutter, 1904}, p. 75).

This is something that Hope Olson, after applying an iteration technique on Cutter’s discourse and interpreting that he meant the name that fits the essence of the concept best, might disagree: «“American literature” does not express an essence: It is a term of convenience. Ask any Canadian» (^{Olson, 2002}, p. 89).

Conclusion

Choosing appropriate headings in both classificationist and classifier processes is crucial since the right heading can help to find books on a topic, using the language of the marginalized group/user to match both descriptions and queries, while the wrong language can cut off access and confine knowledge of these books to oblivion. In a practical way, classifiers will to some degree be able to make amendments through their application of the system and intentionally biased interpretations of the content of a book. The sum of these decisions will affect the way people reflected in those groups will be represented in library catalogs worldwide, the way those people will look for themselves, misunderstand themselves, or feel totally ignored by the system/their society, and the way different societies around the world will perceive those representations through online access. While the minoritization vs. universalization issue presents no consensus in the literature on the representation of unprivileged groups in knowledge organization systems, our study shows that the minoritization view is needed to conduct intersectional studies. The consequences of intersectionality have wide-ranging consequences and cannot be dismissed as hidden in a library catalog so the results suggest a responsible naming to represent the facets and the groups. While more research is needed to mitigate the dangers of essentialism in this regard, some techniques such as strategic essentialism, crossroads or exuzilhada are have been presented as possibilities.