FAIR, CLEAR, and Causal: A semantic framework for ecological knowledge synthesis

What this was about

Lars Vogt presented a purely conceptual framework, developed in the EcoWeaver initiative, for representing causal ecological knowledge with Semantic Units. Using Darwin's naturalisation conundrum (phylogenetic relatedness can raise or lower establishment success), he modelled causal relations as binary relations between change processes. He introduced universal causal statement units with their own identifiers, provenance and asserted/not-asserted status, so contradictory claims can coexist without logical inconsistency. Paths, networks and causal perspectives get addresses in a 'semantic grid', like map addresses. Evidence, cases (linked via a negatable 'satisfies' statement), conditions, measurement methods for unobservable variables and formal model readings can be pinned to them.

Why it matters. Ecological synthesis produces causal claims faster than they are consolidated. A representation that keeps contradictory claims, their evidence and their contexts queryable could support evidence-gap mapping and causal modelling.

In the room

• Running example: Darwin's naturalisation conundrum, with pre-adaptation and limiting-similarity paths giving opposite effects from the same cause.
• Causal relationships are modelled as binary relations between processes of change in qualities or dispositions and their bearers.
• Three requirements: support intervention reasoning (AAA principle of actionable knowledge), distinguish general claims from particular cases, and represent change processes and methods for quantifying unobservable qualities.
• Many needed statement types are not supported by OWL semantics, hence Semantic Units: one identifier per semantically meaningful unit, technology-agnostic, with dynamic labels/graphs for CLEAR human readability, and negation in the A-box without complex class axioms.
• Universal causal statement units carry provenance, literature source and extraction method, and whether the proposition is asserted, so contradictory claims can coexist.
• Compound units give identifiers to paths, networks and causal perspectives (including chains, forks and colliders), forming a semantic grid comparable to OpenStreetMap addresses.
• Cases are linked to universal claims through a 'satisfies' statement that can be negated to record failures, making failures easy to query. All units are versioned and citable.

Notable moments