Coordinating dissent as an alternative to consensus classification: insights from systematics for bio-ontologies

Decentralized but Globally Coordinated Biodiversity Data

“…As we use the term here, a computer ontology provides standardized metadata vocabularies organized into formal logical relationships that allow users to categorize and reason about data in a machine-readable format. Many of the largest bio-ontologies, such as the Gene Ontology, are members of the Open Biomedical Ontologies (OBO) Foundry, which historically has been associated with a metaphysically realist approach to selecting and defining metadata vocabularies (Smith et al, 2007;Sterner et al, 2020).…”

“…As an example, we have shown how multiple locally and simultaneously applied meanings of taxonomic names for endangered Southeastern United States orchids in the Cleistes/Cleistesiopsis complex will produce misleading ecological and conservation inferences if aggregated under one centralized classification (Franz et al, 2016b; see also Peterson and Navarro Sigüenza, 1999). These issues with conflicting taxonomic labels is just one example in a broader debate about the desirability of consensus on a single classification system for a domain of data and appropriate governance strategies for managing advances and disagreements about metadata categories, especially when their definitions are justified by empirical hypotheses and evidence (Sterner et al, 2020).…”

“…Crucially, a decentralized approach permits multiple metadata systems to coexist when important to the functioning of relevant communities of practice while ensuring sufficient resources exist to allow efficient and accurate translation across locally variable metadata systems. Decentralization therefore relies on a second-order consensus about how to coordinate across local variation among communities of practice rather than a first-order consensus about a single global best option for all such communities (Sterner et al, 2020). Our conceptual model therefore extends a core tenet of philosophical pluralism into the context of knowledge commons: "What is crucial for your ability to communicate with me-to convey to me information about your beliefs, plans, or values-is not that we have a commonality of beliefs or ideas, and so stand in a consensus of some sort.…”

“…More broadly, biodiversity loss is also a good example of a "wicked" problem: "one cannot understand the problem without knowing about its context; one cannot meaningfully search for information without the orientation of a solution concept; one cannot first understand, then solve" (Rittel and Webber, 1973, p. 162). Incorporating the possibility of dissent and local customization into the fundamental architecture of the data ecosystem may therefore prove crucial to harnessing the data revolution to address biodiversity loss (Sterner et al, 2020).…”

“…The historically dominant paradigm, by contrast, has depended upon a stable or coherently evolving consensus among multiple scientific communities about the best way to classify natural phenomena (Godfray, 2002;Smith et al, 2007;de Jong et al, 2015;Ruggiero et al, 2015;Sterner et al, 2020). Following this best-consensus model, first-generation big biodiversity data projects have generally prioritized scale and comprehensiveness, launching large initiatives to centralize access through national or global web portals (Wieczorek et al, 2012).…”