Increasingly, universities and research institutes have turned their attention to preserving their institutional memory and scientific–cultural heritage. In this context, university libraries play a strategic role that goes beyond storing and conserving books; they also engage in mediating, providing access to, and preserving various informational resources, including physical and digital documents and historical scientific objects.

Among these resources, teaching and research instruments stand out as representations of the material culture of science, reflecting pedagogical and investigative practices from different periods. The Professor Manoel Lopes de Siqueira Library, a graduate studies library within the Department of Physics at the Federal University of Minas Gerais (UFMG), recognised this relevance and developed an initiative for the identification, preservation, and exhibition of historical scientific objects used in the department’s laboratories and teaching activities. Comprising approximately 50 items, such as temperature controllers, spectrometers, magnifying glasses, and variable resistors, this collection includes pieces dating from the 1940s to the 1990s, which are now considered obsolete due to scientific and technological advancements.

Although they are no longer used in experimental and educational activities, these objects constitute valuable sources for research in the history of science, physics education, and the preservation of academic heritage. Their disappearance could represent an irreparable loss to the memory of scientific and educational practices. The structured documentation of these objects through description and indexing processes is a fundamental strategy to ensure they are safeguarded and to communicate their value and history to future generations (Silva et al., 2016).

In this context, the library faces several challenges regarding this type of collection. The first concerns organisation and description, as there is a lack of specific standards in the literature that address scientific instruments. These objects do not belong to the group of informational resources with traditional bibliographic description formats and, thus, require cataloguing with specific fields. Another issue is the absence of controlled vocabularies since the description requires technical terms specific to physics and the history of science. Furthermore, many instruments have incomplete documentation with unknown origins or functions, which makes cataloguing difficult.

The second challenge relates to preservation and conservation, as the collection is composed of various materials, many of which are fragile and require specific care. Natural degradation over time is also a factor, as the objects suffer from natural wear and tear. Furthermore, climate and humidity play a role since the library lacks a climate-controlled environment and is located in a tropical climate with high temperature variation and humidity; these conditions accelerate the deterioration process of the instruments.

The third challenge is related to management. Few specialised staff members, such as conservators and museologists, are available for these kinds of instruments. Finally, funding is absent, with scarce resources for inventory, restoration, and outreach activities.

Considering that librarianship has historically focused its efforts on the representation of textual resources, this article presents a proposal for a descriptive record form adapted for historical scientific objects, developed by the Library of the Department of Physics at UFMG. The initiative is based on the analysis of experiences from other institutions with similar collections and on the guidelines found in the scientific literature on metadata and cultural heritage items’ descriptions. Although the structure of the form is based on the VRA Core standard and the descriptive models used by Harvard and the National Museum of Natural Sciences (Museo Nacional de Ciencias Naturales – MNCN), this article discusses the reasons behind these choices and the limitations encountered. Additionally, it explores the possibility of adapting or combining these models with other standards that are more specific to the scientific context. This article aims to support other institutions with similar collections in replicating and adapting the form to their contexts.

Literature Review

Information and communication technologies have transformed various sectors of society, including libraries, which have been able to make their catalogues and collections available online. However, for this to happen, it is necessary to use standards for the description of information resources. Among these standards, metadata plays a key role in enabling the creation, organisation, description, identification, and access to information resources (Ferreira et al., 2018). The use of metadata is grounded in the traditional activity of cataloguing in libraries, aiming to describe informational resources to make them unique among other existing materials and provide various ways of access so that the user can retrieve them in an informational system environment, whether conventional or digital, as noted by Formenton et al. (2018).

Among the functions of metadata, the following can be highlighted (Iannela & Waugh, 1997): presenting the meaning of data, allowing the data to be located, identifying if the data meets the need, preventing certain inappropriate uses, retrieving and accessing copies of the data, assisting in the interpretation of the data, identifying the conditions of use, presenting information on the history and trajectory of the data, highlighting the author or owner of the data, showing connections to other resources, and controlling and managing the data.

Among the types of metadata, five categories can be highlighted (Senso & de la Rosa Piñero, 2003):

1. Administrative: Administrative metadata is used for the management and administration of information resources, such as information acquisition and version control.
2. Descriptive: This type of metadata is used to represent informational resources, such as catalogue records and user annotations.
3. Preservation: Preservation metadata focuses on preserving informational resources. Examples include metadata that indicates the conditions for using physical resources and the actions taken to preserve physical or digital versions of a resource.
4. Technical: Technical metadata is related to how the systems or metadata behave. Examples include resource format and hardware and software documentation.
5. Usage: This type of metadata is related to the level and type of use made of the resources. Examples include information about resource versions and content reuse.

Currently, there are various metadata standards. Examples include Dublin Core, EAD, METS, MODS, PREMIS, and VRA Core. A metadata standard can be defined as a set of specifications for describing informational content. These standards are considered formal structures, also known as metadata schemes, which are sets of defined elements aimed at describing a specific type of informational resource. From this perspective, a set of metadata elements consists of two fundamental aspects. The first aspect, semantics, refers to the meanings and definitions of the elements and their possible refinements. The second aspect is content, which refers to the guidelines on what values should be assigned to these elements and how this assignment should be done. Therefore, for each element defined in a standard, guidelines are established on how content should be created or recorded, such as by identifying the main title of a resource and date and time formats and establishing what values can be used, whether through controlled vocabularies or data extraction directly from the document (Formenton et al., 2018).

Dublin Core is a metadata standard that, in its simple version, consists of 15 elements aimed at facilitating the discovery of electronic resources. Originally, this standard was designed for the description of web resources by their creators; however, over time, it has attracted the attention of formal communities such as museums and libraries (Weibel, 2005). The 15 elements of the simple version are described as follows by Weibel (2005):

1. Title: The name given to the resource by the creator or publisher.
2. Creator: The person(s) or organisation(s) primarily responsible for creating the intellectual content of the resource.
3. Subject: The topic of the resource, namely keywords or phrases that describe the subject or content of the resource, including controlled vocabularies or classification schemes.
4. Description: A textual description of the content of the resource, including abstracts for document-like objects or content descriptions for visual resources.
5. Publisher: The entity responsible for making the resource available in its present form, such as a publisher, university department, or corporate entity.
6. Contributor: Person(s) or organisation(s) who made significant intellectual contributions to the resource but whose contribution is secondary to those in the creator element (e.g. editors, transcribers, and illustrators).
7. Date: The date the resource was made available in its present form.
8. Type: The category of the resource, such as homepage, novel, poem, working paper, technical report, essay, or dictionary. It is expected that the type element will be chosen from an enumerated list of types.
9. Format: The data representation of the resource, such as text/html, ASCII, Postscript file, executable application, or JPEG image.
10. Identifier: A string or number used to uniquely identify the resource. Examples include URLs and URNs (when implemented). Other globally unique identifiers, such as ISBNs, are also candidates for this element.
11. Source: The work, either print or electronic, from which this resource is derived, if applicable.
12. Language: The language(s) of the intellectual content of the resource.
13. Relation: Relationship to other resources. This element is intended to express relationships among resources that have formal relationships with others but exist as discrete resources themselves.
14. Coverage: The spatial and temporal characteristic of the described resource.
15. Rights: A link to a copyright notice, a right-management statement, or a service that would provide such information dynamically.

The qualified version of Dublin Core includes three additional elements, namely audience, provenance, and rights holders (Formenton et al., 2018).

The Encoded Archival Description (EAD) schema was developed in 1993 through a project by the University of California, Berkeley (Formenton & Gracioso, 2022). It is an XML-based standard used to encode archival finding aids. Currently, the EAD is maintained by the Technical Subcommittee for Encoded Archival Standards of the Society of American Archivists, in collaboration with the US Library of Congress (SAA Technical Subcommittee for Encoded Archival Standards, 2023). The EAD is currently in version EAD3 and contains various elements and attributes organised into sections that maintain the structure of archival description. This schema allows for contextualised descriptions of resources, helping users categorise and locate the information (Barbedo et al., 2007).

The METS standard was developed from a project called Making of America II (MOA2), which resulted in a metadata encoding system for textual and image materials. This system used XML to establish a standard for organising descriptive, administrative, and structural metadata. METS allows for the management of digital objects in repositories and the transfer of data between institutions or users (Formenton et al., 2018).

The Metadata Object Description Schema (MODS) was launched in 2002 by the Library of Congress. It is an XML-based bibliographic metadata standard intended for digital library objects. This schema uses textual elements and is, therefore, more accessible (Formenton et al., 2018). Currently, MODS is in version 3.8 and can be used for various purposes, especially in libraries (Library of Congress, 2022). From the perspective of digital preservation, three relevant elements include source information, which provides the provenance of the digital object; related item, which aims to analyse connections to other resources; and access conditions, which records usage restrictions or permissions. These fields help ensure the authenticity, integrity, and traceability of digital objects (Formenton et al., 2018).

This study considers scientific instruments as three-dimensional images, and to prepare the description sheet, it uses the standards applied for image description. Among these standards is the VRA Core, a data standard used to describe works of visual culture. This metadata standard facilitates a common language among those who manage artistic assets and establishes a common foundation for those who observe art (Martynovich, 2024). The standard is hosted by the Network Development and MARC Standards Office of the Library of Congress in partnership with the Visual Resources Association (Visual Resources Association, n.d.). The VRA Core can be viewed as an extension of the Dublin Core but with a focus on image resources, particularly works of art. This standard distinguishes between original works (e.g. paintings, sculptures, and architectural works) and reproductions of artworks (e.g. slides and digital photographs). The latest version of VRA Core is 4.0 and is internationally recognised as the only metadata standard created specifically to describe images and cultural objects (Ferreira & Santos, 2013; Mandal, 2018).

In VRA Core, three different types of description are presented: Work, Image, and Collection. The ‘Work’ category covers events or objects of cultural production, for example, buildings, vases, paintings, and performances. The ‘Image’ category refers to the visual representation of the object or event, such as a digital image of a work of art or a photograph of a building. The ‘Collection’ category allows the cataloguing of groups of information resources, including images. In the context of VRA Core, ‘elements’ correspond to metadata and can be compared to fields in a database. ‘Sub-elements’ are elements that have a hierarchical relationship with the main elements. Finally, ‘attributes’ qualify or establish relationships between the metadata of different elements or sub-elements (Ferreira et al., 2018).

VRA Core 4.0 is expressed in XML, which allows it to be used in different software. Additionally, it supports the use of attributes and the hierarchical structuring of metadata (Ferreira & Santos, 2013). These elements are detailed in Table 1.

Concerning historical scientific objects, an example of an institution with a collection is Harvard University. This institution makes the collection available online to provide greater access to these objects and records. The latest generation includes the electronic component called Waywiser, which allows online visitors to browse the collection and search for information about the items (Harvard University, n.d.). The collection began with the initiative of David Pingree Wheatland, who observed that the instruments discarded on the premises of the university’s Physics Department were objects he had seen in photographs from his collection of rare books. Over time, Wheatland formed this collection of historical objects that have great scientific relevance (Harvard University, n.d.). The descriptive record from Harvard for scientific objects originated from the work developed by university museums and academic libraries at Harvard University. The main objective was to create a standardised description for this type of informational resource, enabling these materials to be preserved and made accessible. The record was developed out of the need to provide detailed descriptions of historical scientific instruments belonging to the collections of the Collection of Historical Scientific Instruments (CHSI) at Harvard University (Harvard University, n.d.). Regarding the description of the collection, research conducted in the university’s online catalogue revealed that the collection uses the description fields presented in Table 2.

Another example of an institution that developed its own cataloguing system is the MNCN in Madrid. The MNCN has a collection of historical scientific instruments consisting of various types of objects dating back to the 18th century, which were a part of the Real Gabinete de Historia Natural. This museum adopted a cataloguing system adapted to the collection, and the description comprises 27 fields that provide information about the identification of the item, namely number, name, location, provenance, current state of conservation, and future maintenance needs. Moreover, this description allows for the management of restorations and the loan service for exhibitions or other institutional needs. The knowledge of the pieces is also extended to the incorporation of bibliographic data, both specific and general for each item, and the corresponding archival references (Osuna et al., 2022). Below, in Table 3, the museum’s description fields are presented.

An important initiative in the management and organisation of scientific instruments was the development of the Thesaurus of Scientific Instruments in the Portuguese Language, which was carried out between 2006 and 2013 by a network of institutions from Portugal and Brazil. This initiative was coordinated by the National Museum of Natural History and Science, part of the University of Lisbon, and by the Museum of Astronomy and Related Sciences (MAST) in Rio de Janeiro, affiliated with the Interuniversity Centre for the History of Science and Technology (Centro Interuniversitário de História das Ciências e da Tecnologia, n.d.). However, the official website of the thesaurus is no longer available for consultation.

Methodology

This section describes the methodological procedures adopted for the development of the descriptive record of the historical scientific instrument collection of the Library of the Department of Physics at UFMG. The research follows a qualitative, applied approach with a descriptive objective, focusing on the organisation and representation of information on scientific objects in the context of a library that houses a diverse collection.

The methodology was structured into five main stages: (1) review of the scientific literature on existing metadata models; (2) analysis of descriptive models used in different institutional and disciplinary contexts; (3) selection of the most suitable descriptive elements for the characteristics and needs of the collection in question; (4) development of a descriptive record proposal adapted to the library’s collection; and (5) application of the descriptive record to instruments in the library.

For the initial stage, three referential models with recognised applications in the description of cultural, scientific, and artistic objects were selected: the VRA Core, which is widely used for describing visual works and objects of historical value; the descriptive record of the Department of the History of Science at Harvard University, which is specifically focused on scientific instruments; and the taxonomic record of the MNCN, which offers a detailed structure for the classification of natural and scientific objects.

The choice of these models was justified by their relevance in handling collections similar to that of the Physics Library, their diverse approaches (visual, scientific, and museological), and the richness of their descriptive elements. The comparison between these standards made it possible to identify convergent and divergent categories with complementary functions, enabling a critical analysis of the most appropriate fields for the reality of the library’s collection.

Table 4 presents a comparative analysis of the three models, highlighting elements with direct correspondence or functional similarities. Based on this analysis, the most relevant fields were selected for the construction of the final descriptive record, aiming to ensure representativeness, standardisation, and usability of the information by library users and the scientific community.

Table 4 shows that some elements presented across the three models differ in terminology and conceptual focus. The comparison between the VRA Core, Harvard Historical Objects, and MNCN systems proved relevant and necessary. The three instruments present distinct approaches: the VRA emphasises stylistic aspects, as it is linked to images and works of art; the Harvard model has a functional and historical perspective; and the MNCN model has a taxonomic approach. Despite these differences, they all share the common goal of documenting, preserving, and contextualising scientific and historical objects. The comparative analysis facilitated identifying the elements present in each model from a perspective of convergences and gaps between the standards, which had to be adapted to the needs of the collection housed in the physics library.

After comparing the standards, the fields that could be used by the Library of the Physics Department at UFMG were identified. The three standards offer relevant and complementary elements for cataloguing and managing informational resources. Each of the selected fields provides a description aligned with the library’s needs, allowing for the organisation, preservation, and access to the collection. Table 5 presents the fields selected and created by the library based on the three chosen models.

For the development of the description model adopted in the Physics Library at UFMG, criteria that encompassed both traditional bibliographic organisation and the specificities of the scientific and instrumental objects in the collection were considered. The choice of fields was made from the perspective of the need for a complete and functional representation of the items, ensuring their identification, historical, technical, and material contextualisation and enabling strategies for preservation and information retrieval.

The field ‘Number in the catalogue’ was included for its importance in the unique identification of each item within the institution, allowing effective asset control and internal tracking. ‘Title/Classification’ serves the basic function of naming and classifying the object, facilitating its location and consultation. The fields ‘Agent’ and ‘Maker’ record, respectively, the intellectual creator and the manufacturer of the object, contributing to the attribution of authorship and the contextualisation of technical production.

The ‘Description’ field provides a general description of the object, while ‘Subject’ situates it within a thematic area, contributing to the organisation of the collection. The ‘Date’ field is relevant for historical studies and technological evolution analysis. The ‘Work type/Function’ field clarifies the original function of the object, essential for its understanding in educational or research contexts.

Considering material aspects, ‘Material’ and ‘Measurements/Dimensions’ fields are fundamental for both the conservation and planning of storage and display. The library is located on the fourth floor, and the weight of a scientific object must be known to ensure the safety and structural integrity of the building. The ‘Rights/Bibliography’ records sources of consultation and any legal restrictions on use or reproduction. ‘Location/City of use’ identifies where the object was used or is stored, providing context for its institutional use.

‘Lead Researcher’ highlights the connection of the object to specific researchers, useful for the history and identification of academic or experimental trajectories. ‘Accessories’ allows for the recording of complementary parts, common in scientific instruments. ‘Acquisition method’ reveals how the item was incorporated into the collection, which can include information regarding the researcher or laboratory that donated it to the library. Moreover, ‘Conservation’ documents the item’s current physical state, which is essential data for curation and preservation.

The association of the object to a larger set or collection is covered by the ‘Work, collection, or image/In collection(s)’ field. ‘Inscriptions’ allows for the recording of marks, texts, or signatures present on the object, elements that often carry valuable information about its origin or use. ‘Keywords’ was included to facilitate thematic indexing and retrieval in digital systems. To describe the keywords, a controlled vocabulary used by the UFMG Library System is applied. Finally, the ‘Notes’ field provides a free space for additional observations, ensuring descriptive flexibility and the inclusion of information that does not fit into formal fields.

This set of fields, therefore, seeks to meet both the demands of an efficient informational system and the care required by the technical and historical nature of the Physics Library’s collection at UFMG, integrating heritage, academic, museological, and library sciences aspects.

Results

This section presents the results of applying the model for describing scientific and historical objects in the Physics Library. As a sample, the repeating theodolite (Figure 1) and the microdensitometer for astronomical plates (Figure 2) were selected and described (Tables 6 and 7).

During the development of the descriptive record for the historical scientific instruments of the Physics Department Library at UFMG, it was possible to research and identify different metadata standards across various contexts. Throughout this experience, it became clear that although the literature presents models with distinct approaches – such as artistic and bibliographic – in practice, institutions often use customised models adapted from existing ones. Significant convergences were identified in creating this record and through bibliographic research and consultation with institutions holding similar types of collections, which can be explored complementarily in each model used. Adapting the categories to the reality of the collection required a critical perspective on the specificities of the holdings, valuing the technical aspects of the objects and their historical and scientific value and the trajectory of the experiment from the perspective of its past use and current presence in the library. Moreover, the process reinforced the relevance of standardised practices in information organisation to ensure the preservation, accessibility, and dissemination of scientific heritage.

Another important insight gained during the preparation of the descriptive record was the recognition of the relevance of interdisciplinarity. Describing the instruments required knowledge of physics, astronomy, museology, and library science, highlighting the need for collaboration between different fields to preserve and contextualise scientific objects.

Finally, the practical application of the model confirmed its potential as an educational tool, as the description of the instruments, including explanations of their function and history, supports the scientific outreach already taking place in the library through guided tours and via social media, connecting the past and present of research.

In summary, this study reinforced the idea that detailed and accessible documentation of historical instruments not only preserves their memory but also enhances their usefulness for teaching, research, and outreach.

Conclusion

This study confirmed the effectiveness of the descriptive model developed for historical scientific objects in the Physics Department Library at the Federal University of Minas Gerais (UFMG). The model was based on a comparative analysis of three metadata standards – VRA Core, the Harvard model, and the MNCN system – and integrated elements from each one to address the specific historical, informational, and material characteristics of the collection.

Applied to objects such as the repeating theodolite and the microdensitometer, the proposed model enabled detailed and contextualised documentation, capturing technical specifications, historical significance, provenance, and educational function. This level of description enhances the organisation, preservation, and dissemination of institutional memory and scientific heritage.

Furthermore, the model is valuable for enhancing information interoperability and retrieval, promoting standardisation in the cataloguing of non-bibliographic materials. It supports the broader objective of integrating scientific cultural heritage into the library and archival systems, offering the potential for reuse in digital repositories, online exhibitions, and educational outreach. In this regard, the inclusion of elements such as controlled vocabularies, acquisition history, conservation status, and researcher attribution enriched the model’s descriptive framework and aligned with best practices in digital curation and museology.

The development process underscored the necessity of interdisciplinary collaboration, drawing on knowledge from library and information science, museology, physics, and the history of science. It also highlighted the institutional challenges that affect the sustainability of such initiatives, including limited resources and the lack of specialist personnel for conservation and heritage management.

In light of these outcomes, the model can serve as a reference for other institutions managing similar collections. Future research should focus on refining and applying the model to broader collections, testing its adaptability across different scientific domains and institutional contexts. Moreover, there is value in exploring its integration into metadata standards, fostering a more unified approach to the preservation and visibility of scientific and academic heritage.

Ultimately, this work affirms that a robust and context-sensitive descriptive framework is essential for the long-term safeguarding, accessibility, and appreciation of historical scientific instruments. This framework not only preserves the material legacy of scientific endeavour but also reinforces the role of libraries as stewards of cultural and scientific heritage.

References

Abstract

This article aims to present a description model for historical physics objects, which was developed for the library of the Physics Department at the Federal University of Minas Gerais (Universidade Federal de Minas Gerais – UFMG), Brazil. The research adopted a qualitative approach with an applied nature and a descriptive objective, focusing on the organisation and representation of information related to scientific objects in the context of a library with a diverse collection. The methodology involved a comparative analysis of three description models: the VRA Core, used to describe visual resources related to art; the description standard developed by the historical scientific objects collection at Harvard University; and the descriptive record from the Museo Nacional de Ciencias Naturales (MNCN) in Madrid, Spain. This study included a detailed analysis of the metadata fields present in the three models, identifying the relevant information from each one. Based on this analysis, the research proposed the adaptation of these models to meet the specific needs of the historical physics objects at UFMG, which reflect the department’s teaching and research, thereby preserving its institutional memory. As a result, a hybrid description model was developed, combining elements from the VRA Core, which emphasises visual description, with those from the Harvard and MNCN models, which highlight the functionality and historical context of the instruments. The research also involved applying the model to an object from the library’s collection. The findings show that its application contributes to a more comprehensive and contextualised description, supporting the preservation and retrieval of the informational object within an information system and assisting with guided tours and the dissemination of the collection through the library’s social media channels. Furthermore, the model can serve as a reference for other libraries or information centres in the organisation and accessibility of scientific objects.

Keywords

Historical physics instruments; Memory preservation; Descriptive record; Metadata; Information representation; Universidade Federal de Minas Gerais