There is some irony in that many of the terms – ‘thesaurus’, ‘taxonomy’, ‘controlled vocabulary’ ‘ontology’ – that are intended to bring order and clarity to our use of language in professional settings are themselves subject to diverse interpretations and application. This is in large part because they are used by a range of people working in different contexts with different purposes. When it comes to standardised vocabularies, there are several areas of application. The question is thus whether (and how) any one tool can accommodate the full range of potential uses or whether it becomes of no value to any in the attempt. To answer this question, we need to understand what the main uses of such a vocabulary are.
First, such vocabularies are essential tools of information management. One of the major roles of Land Portal is to make information about land governance readily accessible. It either holds, or links to off its website, tens of thousands of documents and datasets. The Land Portal, along with any other content platform, requires a system to describe all of this material in ways that its wide range of users can find and use the material of most relevance to them. Historically there have been a number of ways of doing this, from the subject-based Dewey Decimal system used in libraries, to many varieties of content management software, both open source and proprietary, to user-driven tagging systems or ‘folksonomies’.
All content curators have such systems, even by default. Because a key mission of the Land Portal is to link knowledge (and the terms used to describe it) across the many disciplines and personal roles that constitute the field of land governance, it needed such a tool to organise its own metadata. It also made sense to develop this tool so that it could be shared with and used by other organisations. Having ‘industry-wide’ standards greatly facilitates the sharing of information in any sector. This is what AGROVOC has been working towards for years in the broad field of rural development and across different specialist sites: LandVoc’ aims to make a more specific contribution around Land Governance. As the adoption of such standards within any field of work is a voluntary process, there needs to be broad agreement as to their validity and, in such a politically charged area as land governance, fairness.
In recent years, the on-line discovery and exchange of information has become ever more pronounced. Digitisation has also led to an explosion in the volume of available material. Making this work for land governance and ensuring that new inequalities or exclusions are not unintended outcomes of the process are also key aims of the Land Portal. One inevitable consequence has been the increasing use of computers to identify and organise information through machine driven processes. To do this, computers need to understand what information they are processing, something which is achieved by producing machine readable metadata (that is data describing the information concerned). To the layperson, this involves the structured use of what are essentially standardised vocabularies, although within the computer science world they are described as ‘ontologies’. Again, using open and standardised methodologies facilitates links with similar knowledge organisation processes in related fields. Thus, both LandVoc and AGROVOC use the Simple Knowledge Organisation System developed as a core component of the semantic web under the auspices of the World Wide Web Consortium. This actually allows more flexibility in how terms and concepts can be mapped and their relationships explained than might be expected. I am not aware of either LandVoc or AGROVOC experiencing any real constraints in mapping their vocabularies into SKOS. However, it does require collaborating with people who come at their work with different perspectives, terminologies and, sometimes, priorities.
Another strand in the specialised use of vocabularies is their use in research. Historically, this has been particularly important in the academic study of the natural sciences and in both the study and practice of professions such as medicine and aviation where misunderstandings can have fatal consequences. This use can reach beyond controlled vocabularies to encompass taxonomies. These are ordered linguistic structures, which assert a particular logical structure for their discipline and require its acceptance and use by participants. LandVoc has never made any claim to be a taxonomy and would not do so. Many of its users, however, are researchers coming from disciplines where they are accustomed to such stringent use of vocabularies and who expect to see a consistent logic applied to how concepts relate to each other. Thus, one academic member of our community of experts thought that the term ‘land governance’ was poorly defined and delimited in relation to its component parts and that ‘this hampers any scholarly use of LandVoc, and blurs the improvement discussion’. The issue is not that of agreeing to a single definition of a ‘standardised vocabulary’ but to aim to create a product which makes sense to and is of value to all its users.
>This is further complicated by another research use of what could be termed linguistic mapping, which is intended not to remove but to explore differences of meaning. A host of participatory methodologies within development studies and geospatial information systems use the ambiguities of language to explore difference and conflict. Such methods are increasingly used to map out areas for inter- and trans-disciplinary research.
What has all this got to do with the development of LandVoc or any other specialist vocabulary? As argued above, for a tool to become a ‘standard’ it has to work for everyone who wants to use it. Our internal evaluation of the continuing development of LandVoc reflected on the difficulty, especially in a world of COVID induced virtual meetings, of fully understanding what attracts people to becoming involved, what their expectations of it are and how best they can contribute. If these are key questions for people offering their time to a Community of Experts, they are as important when thinking of the range of users in general. This blog has attempted to track the starting points and requirements of those who are interested in using LandVoc. It suggests a clear need to understand the different strands of background interest and motivation in a development project of this nature and to ensure there is sufficient discussion and diversity of input to meet it. It thus addresses some of the key human issues in building a successful project. Of course, difference and diversity are often embedded in language. Whether and how they can be accommodated in any useful way within a standardised vocabulary is explored in the blog post, “The continuing development of LandVoc: Becoming a linked, open multidisciplinary space.”