Thesauri and controlled vocabularies

Although thesauri specializing in cultural objects have existed for a long time, it was in the early 1980s that more rigorous and specific standardization began to be developed within the field of visual arts, architecture, new media and the performing arts. Described in a very generic way, a thesaurus is a compendium of terms or phrases that, unlike a dictionary, does not provide definitions but rather determines and restricts the uses to be made of each of them in a given context. Thesauri, therefore, fulfill multiple functions, among them, to establish authorized voices when using polysemic or even antagonistic expressions such as: works of art, document, artifact, Modernism, Baroque style, etc.

They also serve a somewhat opposing function: to link different words that refer to the same concept. Moreover, a thesaurus identifies spelling variations and syntactic errors common in the writing of a word to facilitate the effectiveness of search engines. Furthermore, it is usually multilingual, so that it may function to link terms that are used in different languages that refer to the same concept. In relation to search engines, thesauri allow the crossing of fields and establish disambiguation criteria, that is, they determine terminological families and the degree of linkage between concepts to facilitate optimal and controlled access to information.

The structure of a thesaurus is based on three elements: equivalence, hierarchy, and relationality between terms. To fulfill their objectives, thesauri operate with what are known as controlled vocabularies, that is, with an authoritative compendium of terms and a framework of reference for their descriptions ^{(Harpring, 2010)}. The ability to control a vocabulary determines the functionality and adaptability of a thesaurus over time. Entering a new term into the body of a thesaurus therefore requires a series of complex and sophisticated institutional, conceptual and informational protocols. If we take for example the term hybridization, we would see that it has a diametrically opposed connotation when used in the field of genetics than when used in that of postcolonial theory. Paraphrasing Mieke Bal, hybridization would be one of those traveling concepts that moved from biology to the field of humanities and, in that migration, reversed the sense of its meaning ^{(Bal, 2009, pp. 38-39)}.

From an archival science point of view, the usefulness of controlled vocabularies in document management and information architecture are unquestionable (for example, for cross-catalogue consultation, to establish criteria for qualitative research or in the construction of semantic webs). However, from a humanistic perspective, or more specifically, when considering the impact that these semantic restrictions have on the processes of research, visualization and knowledge construction, the question of validation policies and the conceptual models that give legitimacy to the system of authority over controlled vocabularies must be kept open to diverse communities of experts: how are such compendiums of terminological authority constructed? Who participates in their conceptualization, translation and management? In the end, what kind of epistemological debts do thesauri have to their predecessing illustrated encyclopedic structures that were based on taxonomies and hierarchically organized content?

My opinion is that beyond their mere practical application, the impact of thesauri in the processes of research, canonization and visualization of fields of knowledge, should be considered as objects worthy of study in and of themselves. Thus, while it may seem redundant, I think it is necessary to establish a kind of metacritique of the metadata. One which would deal, not only with the epistemic value of how far one set information accurately accounts for other pieces of information, but also, with the historical and institutional processes through which these sets of information acquire meaning, and are transformed and visualized over time. Following from the earlier example of the concept of hybridization, it would be of great informative value its use in the field of humanities and its disambiguation from that of some medical sciences, but also the very process of its translation, assimilation and legitimation that has been granted to it by its respective community of experts when they accepted it into the field of critical theory and cultural analysis. In other words, the historical evolution behind the meaning of terms constitutes an especially important degree of information that must be preserved as part of a specialized thesaurus.

From what has been said so far, it is easy to recognize that the epistemological dimension of thesauri pose a series of problems to the effective public service given by the institutions that safeguard our archives and even, to the capacity of agency of the communities of experts who give them meaning, order and design the visibility of knowledge. If there is therefore an urgent question to ask the digital humanities, it is precisely to what extent we, the community of humanities and social scientists, with our restricted documentation skills, are trained to intervene and transform the informational architecture and archival concepts that form the basis of our research and heuristic enquiries. In other words, to what extent is it possible, or even desirable, for humanists to stop being mere end-users of such architectures and to become involved in their conceptualization, modeling and permanent transformation ^{(Benardou, Champion, Dallas & Hughes, 2017)}. What the collective and non-hierarchical construction of thesauri ultimately proposes is the idea of the archiving and cataloguing processes as a place of negotiation, as a community or public sphere.

Folksonomy: the description of documents like agora

As a collective information-gathering strategy, folksonomy is now common practice, despite the fact that the term has not been sufficiently disseminated.² The free assembly of tags or categories in a social network is a type of folksonomy; that is why it is sometimes known as “social tagging”. In essence, folksonomy consists of the uncontrolled, deterritorialized and to a certain extent unhierarchical creation of metadata. Contrary to taxonomy, it is not constituted vertically by a recognized and legitimated community of experts, but by a variety of users who casually and freely compile categories around a given subject. Their degree of authority is low, as it is generally difficult for users to be consistent in constructing categories and families of concepts. Their terminological edifice is not therefore, a semantically controlled construction like that of a traditional thesaurus but rather, an intermittent flow of fragments, superimpositions, contrapositions, linkages, and repetitions.³

It is not possible for me to raise here the technical problems involved in establishing interoperable models between controlled and uncontrolled vocabularies; nor can I discuss the utopian dimension on which this supposed form of informational democracy rests, nor even, the relevance of the folk root (communities, common people) as an alternative to the taxis root (order, arrangement).⁴ Suffice it to say that, for the moment, its emergence and massification raises the following axioms: the less hierarchical and less controlled the vocabularies are, the less authorized and specialized the thesauri are; the greater the standardization and terminological control of a thesaurus, the less diverse, plural and multidirectional the concepts are.

If we think about the political and epistemological dimension of categories such as race, Colonialism, Holocaust, etc., the immediate question is how we can politically manage the consequences of the strategic or temporal lack of control of specialized vocabularies in order to intervene in their conceptual taxonomy. Although we do not have a single answer to this question, the fact is that the challenges of folksonomy force us to rethink the technical and disciplinary division between the informational sciences and the new digitalization of the humanities; in other words, to question the lack of dialogue between those who construct the informational edifice (the cubes or PRIMARY STRUCTURES⁵ of knowledge) and those who use this edifice to carry out a critique of the social world (the political-social performativity of the archive and the very purpose of the public service of museums).

Translating these issues into the specific context of a contemporary art museum immediately raises the following questions: to what extent are the standardized descriptive tools used by cataloguers able to include the participation of other expert communities, such as researchers, curators or semi-specialized audiences, in the processes of describing cultural objects? And, conversely, to what extent are researchers, art critics, curators and artists willing to actively intervene in the design and adaptation of the conceptual tools used by cataloguers? On the first question it is easy to recognize that, over the past 20 years, the archival community has developed a number of specific tools for the cataloguing of art works and cultural objects, most of which have the capacity to include the voice of other communities of experts in descriptive practice and metadata construction (^{Baca, 2006}; ^{Baca & O’Keefe, 2008}; ^{Harpring, 2010}; ^{Lanzi, 2010} & Tillett, 2002).

Such would be the case of CDWA (Categories for Description of Works of Art), which, in the words of Elisa Lanzi, “was formulated for the needs of those who record, maintain, and retrieve information about art, architecture, and material culture, focusing on academic researchers and scholars” (Lanzi, 2010, p. 5672). In its latest versions, the CDWA includes a category called “critical responses”, designed to record “critical opinions about a specific work by artists, art historians, art critics, art dealers, sellers and buyers, public officials, and the general public” (Lanzi, 2010, p. 5673). At the same time, as Murtha Baca has pointed out, from conceptual models such as CIDOC-CRM (Conceptual Reference Model) it is feasible to put into operation social tagging tools, that is, multidirectional systems of a collective construction of tags. In her own words:

Incorporating input from curators, experts, and other specialists is an area that institutions must foster if they want to generate rich and accurate descriptions of the non-bibliographic works in their collections. Information from non-cataloging experts could be collected routinely if there were effective methods of communication and collaboration between cataloguers and curators. Social tagging by experts (that is, the inclusion of keywords, names and topics by experts who are not part of the official cataloging unit of the institution) can be, at the same time, an effective method to enrich the records descriptive metadata ^{(Baca & O’Keefe, 2008, p. 3)}.

Thesaurus epistemic authority

Since the 1980s, the Getty Foundation has been developing various specialized vocabularies. One of them being the Art & Architecture Thesaurus (ATT), a vocabulary of authorized terms specialized in art, architecture, decorative arts, material culture and archival material. The ATT currently has about 375,000 terms structured hierarchically from concepts organized through some 60,000 entries or records called subjects. The ATT is a poly-hierarchical thesaurus, that is, it has different levels of internal hierarchy, so that a term may be related to one or more conceptual families. One of the most important hierarchies of the ATT is linguistics, meaning that it is a multilingual vocabulary in which English is the dominant language:

In a completely dominant vocabulary, all languages are treated equally, with none serving as a so-called dominant language. However, in practical applications it is often necessary to treat one language as the default dominant language, particularly when the vocabulary is rich and complex. An example is the ATT, in which each concept record includes over one hundred fields or data elements in addition to the term itself […] For the ATT, English is the dominant language, although terms and scope notes may be in multiple languages. In addition, if every term in the original source language has not been assigned equivalent in all other target languages, the status of the other languages is not equal to that of the source language, and they are known as secondary languages ^{(Harpring, 2010, p. 92)}.

However, despite its status as a secondary language, in the last decade Spanish has begun to expand its presence both in the public sphere and in the documentary and computer environments of American institutions. That is why, as Chilean researcher Cristián Gómez-Moya recalls, in the second half of the 1990s the Getty Vocabulary Program and the Dirección de Bibliotecas, Archivos y Museos (Dibam), part of Chile’s Center for the Centro de Documentación de Bienes Patrimoniales (CDBP), began the process of translating the ATT into Spanish ^{(Gómez-Moya, 2010)}. The result of 10 years of work was the creation of the Tesauro de Arte & Arquitectura (TAA), which consists not only of the translation of terms into Spanish, but also in the creation of equivalences and preferred terms for the different uses of Spanish terms throughout the world. This is important for our analysis since, to a large extent, the aim is to extend the linguistic hierarchy of the dominant language (English) and the dominant vocabulary (ATT) to the different local registers of Spanish without losing the interoperability of the translation. If I have expanded on the general foundations of the Art & Architecture Thesaurus and its derivative, the Tesauro de Arte & Arquitectura (TAA),⁶ it is because both operate like the “archival grammar” of the ICAA Documents Project.

In January 2012, the ICCA at the MFAH first made available to the public part of its Documents Project, an ambitious project to digitalize documents of Latin American and Latino art of the 20th century.⁷ On the ICCA website itself, this project is defined as:

the result of a decade-long, multimillion-dollar initiative to identify and retrieve thousands of primary and critical texts […] by notable Latin American and Latino artists, critics, curators, and others who have played an important role in the development of the art produced along this cultural axis […] the Documents Project addresses the lag in the field of Latin American art history, research, and teaching. The monumental, long-term impact facilitates the pursuit of new knowledge in the field and establishes a legitimate area of scholarship in U.S. universities ^{(Ramírez, 2012, pp. 27-32)}.

To carry out this project, the ICAA -founded in 2001 and directed by Mari Carmen Ramírez -has worked for 10 years with more than 100 renowned researchers as well as numerous educational and cultural institutions located throughout the continent. It can be concluded that we are facing one of the most solid, most ambitious and best coordinated documentary projects linked to Latin American and Latino art developed so far.

However, given its documentary architecture, its geographical framework and its aspiration to create a new area of study in American universities, it seems to me that it is necessary to go beyond the mere recognition of its importance and scope as a tool for research. Because it is already in the second phase of its expansion, and its consolidation, I believe that the impact of this type of project should be analyzed from various perspectives, including considering what epistemological consequences its role as an agent for the standardization of categories related to art in the region, will have. In other words, a critical reading of the metadescriptive authority on which projects of this type are structured is needed.

ICAA Documents Project has two main platforms: a digital archive of documents (Documents of 20th Century Latin American and Latino Art) and a collection of thematically structured documentary collections (Critical Documents of 20th-Century Latin American and Latino Art). The interesting thing is that both platforms are structured from the same set of metadata; that is, one and the other are the result of the standardization of a series of categories (titles, names, authors, subjects, places) and of terms-themes authorized by an editorial committee. The function of these metadescriptors is to structure not only the semantic architecture of digital search engines but also the editorial architecture of publications.⁸ Hence, these are understood as the documentary extension of the digital archive; each document translated into English and published in one of the volumes has both an editorial and digital identifier (digital archive number). As can be seen, this project has been structured with the explicit aim, on the one hand, of putting into circulation a series of authorized voices and, on the other, of becoming a sort of Spanish-language terminology laboratory for the Art & Architecture Thesaurus.

On the other hand, the sheer magnitude of the project entails a series of institutional and epistemological complexities related both to the language policy of the metadata and to the control of the metadescriptive authority of the vocabulary. An example of the former would be the hierarchical nature of the ATT’s multilingual internal structure, which largely conditions the intercultural, decentralized and non-hierarchical claims of the ICAA.⁹ As far as the metadescriptive authority is concerned -from my point of view, the most controversial aspect of the project- the main problem lies in the inevitable process of canonization of the editorial and descriptive categories that support the documentary architecture of the archive. If, on the one hand, access to the digital archive is free and non-hierarchical, then the printed volumes tend to set the thematic categories canonically. Aware of this situation, the first launch of the website through which the digital archive is accessed, coincided with the launch of the first compendium of texts, which is symptomatically entitled Resisting Categories: Latin American and/or Latino?, in allusion to the weight that categories have, mainly when they are disseminated throughout the continent ^{(Ramírez, 2012)}.

However, in view of the inevitable canonization of a series of categories on Latin American and Latino art that this project will entail in the medium term, we need to ask ourselves to what extent it is sufficient to propose a strictly discursive (editorial) RESISTANCE when trying to counteract the standardization of metadescriptive categories, that is, categories that extend their authority through the digital humanities and technologies involved in documentary information management. Moreover, it is essential to question even the effectiveness of a type of resistance that is proposed from within a project that explicitly aims to be the documentary and conceptual foundation of a field of study legitimately recognized in the American academic context ^{(Barriendos, 2009, pp. 91-104)}.

Although it is not time to give definitive answers to these questions, I tend to believe that while the canonization of categories will be inevitable, the truth is that there are various territories not only where the standardization of categories can be DISCURSIVELY RESISTED but also where one can POLITICALLY INTERVENE in the terminological architectures on which our archives are ordered. I am referring to the disruption of the METADESCRIPTIVE AUTHORITY of controlled vocabularies by creating communities of sense that appropriate and diversify the public use of catalographic description. However, as I have pointed out before, just because they can get out of control does not mean that it is easy or that their political and epistemic consequences can be entirely predictable. In fact, carrying out such a type of intervention requires a high degree of archival activism in a field still largely unexplored by institutional critique.

Conceptual art as a disputed catalographic territory

To date, neither the Art & Architecture Thesaurus nor its Spanish version, the Tesauro de Arte & Arquitectura, offer entries that allow for an adequate contextualization of the differences between the categories of “conceptual art” and “conceptualist practices”. This means that when describing, for example, the documentation of a collective experience of aesthetic-political rupture such as Tucumán Arde, one is forced to enter the category of “conceptual art” and correlate it with others, such as “political art”, “Argentine art”, etc.

The case of Tucumán Arde is relevant not only because of its place in the history of Latin American art, but also because of the problems that its institutionalization has entailed in terms of museology and cataloguing. As we know, Tucumán Arde is a name that acquired a set of aesthetic-political interventions and direct actions of semiotic disruption in Argentina towards the end of the 1960s. Broadly speaking, the experience consisted of three stages. First, an artists’ collective traveled to the Tucumán sugar region to document the marginality and poverty that had been generated by Argentina’s economic developmentalist ideology. At the same that a press release was being issued at the Museum of Fine Arts, a series of interventions in public spaces began to appear in the cities of Rosario and Santa Fe, consisting of graffiti with the words Tucumán Arde and the pasting of posters announcing the holding of the First Biennial of Avant-garde Art. Then, the collective carried out an action-exhibition at the Confederación General del Trabajo (CGT) in Rosario, in which it showed through photographs, banners and performances, the conditions of exploitation that were being experienced in Tucumán. Given its radical nature and semiotic strategy, this action was blocked and censored by the police, who prevented the third phase of the project to be concluded in its entirety in Buenos Aires as intended.

From the early 1970s, European and American art historians took Tucumán Arde as the quintessential model of Latin American conceptual art, defining it as the antithesis of the typical Anglo-Saxon analytical, cold, rational and apolitical conceptualism, while the Latin American one was described as ideological-political conceptualism. The entry of archival materials related to Tucumán Arde into the collections of major international museums in recent decades reopened the debate around these antagonistic and stereotypical characterizations, raising questions not only about the collection of global art but also about the cataloguing systems and the most appropriate terminologies to be used when speaking about the different conceptualisms and their historical disputes.

In 2015, the Red Conceptualismos del Sur (Southern Conceptualisms Network) -to which Graciela Carnevale, a member of the collective that carried out the original Tucumán Arde art protests, belongs- published Desinventario. Esquirlas de Tucumán Arde en el archivo de Graciela Carnevale (Un-inventory: Splinters of Tucumán Arde in Graciela Carnevale’s archive) with the intention of analyzing the potential problems involved in referring to the Tucumán Arde archive either as a Latin American model of conceptual art or as a concrete experience of what the book defines as Latin American conceptualist practices ^{(Carnevale, Expósito, Mesquita & Vindel, 2015)}. The publication is interesting for my argument insofar as it invites an analysis of the case of Tucumán Arde from the point of view of the cataloguing protocols and the archive policies of the art museum. In controversial cases such as this one, we are faced with the dilemma of deciding which terminology to use to most effectively highlight the aesthetic-political conjunctures that have defined the documentary and heritage history of art documentation.

Without claiming to offer a single solution for the different possible cases, an alternative would be, in the event that the CDWA model is being used, to include a commentary in the Critical Responses category clarifying the distinction between “conceptual art” and “conceptualisms”, and framing the historical-political context of their disambiguation. However, given that we already have an important body of documents that could easily justify the relevance and authority of alternative categories such as “conceptualisms”, “conceptualist practices”, “non-objectualisms”, etc., it is also feasible to think about granting them DESCRIPTIVE AUTHORITY within the controlled vocabulary. In fact, the editorial metadescriptors of the ICAA project include both “conceptualisms” and “non-objective art”. This is undoubtedly highly suggestive, as it opens up a whole space of theoretical production; however, it also entails two different types of problems that have to be discussed in detail.

On the one hand, as I have already said, the hierarchical structure of thesauri is designed precisely to resist the instability and lack of control of their terminology, which implies a series of technical and conceptual aspects that need to be resolved taking into account the possibilities offered by folksonomy (the creation of diversified communities of experts, the permanent collection of critical responses, the revision of vocabulary entries, intervention in linguistic hierarchies, etc.). On the other, the very idea of equalizing the degree of authority of non-canonical categories so that they operate in the cataloguing process means, to some degree, the loss of their destabilizing function and the beginning of their canonization.

In relation to the latter point (the imminent canonization of a series of categories that until now had operated as counter-discourses), it is worthwhile raising the possibility of granting them some kind of alternative (satellite or tactical) authority with the intention that these type of semicanonical categories can really preserve the critical and disruptive force that they had before they were integrated into the controlled vocabulary as authorized entries. However, such an idea entails, in turn, a number of potential problems that must be taken into account. If we were to apply, for example, the category “conceptualisms” as an ALTERNATIVE AUTHORITY to the category “conceptual art”, we would have to assess at least three scenarios.

A first potential danger would be that “conceptualisms” could be used as a manifestation of "conceptual art" that is, as one of the forms in which art as an idea or concept is materialized (which would leave the canonical basis on which the category “conceptual art” is currently built, intact). Secondly, “conceptualisms” could be understood as the Latin American expression of “conceptual art”, that is, as “Latin American conceptual art” (which would not only leave its terminological authority intact, but would also create a false familiarity between totally dissimilar and even opposing conceptual experiences). A third potential pitfall would be that “conceptualisms” could be understood exclusively as the authorized meaning of the Spanish translation of “conceptual art” that is, as the preferred term to refer to such category in the Spanish language (which would underline the character of Spanish as a secondary language in terms of metadescriptive authority).

Conclusions

As can be deduced from the above, the use of controlled vocabularies in art museums exposes a series of theoretical and practical problems obliging the active and involved participation of different COMMUNITIES OF SENSE, which must discuss the value of FOLKSONOMY and of authorized voices when including new terminologies in the cataloguing processes. Therefore, in order to consider the catalographic viability of categories such as “conceptualisms” we need to first ask ourselves: to what extent it is desirable to claim their descriptive authority in a given documentary context and above all, what the consequences would be of doing so in one way or another.

If we accept that museum archives are places for reflection, negotiation, identification and a construction of the common; that is, if we are willing to consider them as instruments for the activation of political life and for the breaking of consensus in giving body and meaning to social memory, we will also have to accept that the analysis of the metadescriptive authority of controlled vocabularies must involve different communities of meaning, including archivists, social scientists, curators, artists, activists and non-expert audiences, even though they do not all speak the same language or have the same objectives in mind.

In this way, the documentary categories used and consequently made problematic by art museums, would become true territories of dispute, and documentary management would be able to overflow towards the creation of social links and communities of sense more in line with the supposed public service role of museums and with the development of digital humanities.