O design do ToGatherUp: a aplicação do Design Participativo no desenvolvimento de uma ferramenta computacional para a Linguística de Corpus

Abstract

This research addresses a central issue in Corpus Linguistics (CL): the dependency between linguists and software designers. In a context where most linguists lack programming skills and either rely on software designers to develop computational tools for their projects or use pre-existing ones (Ardito et al., 2012), this dependency can give rise to power dynamics and expertise conflicts, often resulting in products that fail to fully meet the real needs of researchers (Gomide, 2020; Smith et al., 2008). In light of this context, this research investigated the design process of ToGatherUp, a computational tool for managing corpus-building projects. The guiding research question was: “How can the end-users of ToGatherUp be included in its design process?”. The thesis defended here argues that the inclusion of codesigners (representatives of end-users) in the ToGatherUp design process makes its future versions more aligned with the real needs of the software's end-users. The general objective of the research was to develop a methodology for the ToGatherUp design project, aiming for the effective and democratic inclusion of its stakeholders in the design process of the software. This methodology was grounded in Participatory Design (PD), a human-centered design approach that values the empathetic understanding and experience of stakeholders, treating them as codesigners on equal footing with designers. PD contrasts with approaches that often neglect the human, social, multidisciplinary, and communicative dimensions of design, resulting in products misaligned with stakeholders' real needs (Ehn, 1989; Krippendorff, 2005). In the methodology, narratives were understood as an essential means for translating practical (first-order) knowledge into propositional (second-order) knowledge during the elicitation of ToGatherUp's requirements. The software prototype was conceived as a "boundary object" (Star, 1989; Wenger, 1998), facilitating communication and knowledge exchange between the different language games (Wittgenstein, 1999) and Communities of Practice (CoPs) of the research participants. The methodology was structured in three phases: Preparatory, Participatory Cycles, and Closing. In the Preparatory Phase, the codesign situation was established, involving the recruitment of codesigners, preparation of the initial prototype of ToGatherUp, creation of its website and Support Center, and development of the Participatory Design Module (PDM), which served both as a data collection instrument and a channel for participant interaction. In the Participatory Cycles Phase, conducted in a single cycle, the methodology was applied, entailing the use of the prototype by the codesigners to create projects, the recording of their narratives about this use, and interactions between them and the software designer for the joint elicitation of requirements. These requirements, after implementation, resulted in the release of four new versions of the ToGatherUp prototype. In the Closing Phase, data analysis and the systematization of results were performed. The research achieved its objective by developing and applying a methodology that enabled the democratic participation of users in the design process. The four published versions of the prototype, shaped by the narratives and dialogue with the codesigners, embody a design that is more aligned with the real practices and needs of ToGatherUp's future users, thus corroborating the research thesis.