OXA

proposal

The document format behind Curvenote, Stencila and eLife. This tab is a proposal to it rather than a reading of it: a Claim block node that OXA does not yet have.

Draft in docs/schema-mapping/oxa-claim-schema.md · we emit .oxa.json per paper

What it requires

OXA represents a scientific document as typed JSON nodes — Heading, Paragraph, Figure, Code, Cite — each block-level, addressable, and carrying children. Citations are typed through CiTO by RFC0005, which is the pattern the Claim node follows.

There is no node for a proposition. A figure is representable; what the figure argues is not.

What we propose

A Claim block node — the same kind of element as a Heading or a Code block — with three additions. This is a proposal to the format, not a reading of it.

A role

Nine values describing the claim's function in the argument, extending Stencila's existing Claim types — which cover theorems and proofs — into empirical science.

A panel anchor

Which figure panel the claim is grounded in. One panel, one computation, one result.

Typed edges to other claims

Following RFC0005's pattern for citations: each edge carries a relation type from CiTO, or one of our subproperty extensions where CiTO has none. 

{
  "type": "Claim",
  "identifier": "hypothesis-distinct-compartmental-roles",
  "role": "hypothesis",
  "children": [{ "type": "Text", "value": "Perisomatic and distal dendritic
      inhibition serve distinct computational roles." }],
  "relations": [
    { "xref": "prediction-distal-dendritic-spike-mechanism",
      "relationType": "claimrel:entails" }
  ]
}

Does it validate

Not yet, and it cannot. The node type does not exist in OXA, so there is no schema to validate against — the exports are what we would emit if the proposal were accepted. Every paper has one, and they round-trip through the Discourse Graphs exporter, which is the only check available.

What it cannot carry

Verification and stance, as with every target here. Neither is in the proposal, because proposing three additions has a chance and proposing five has less — but both are gaps this corpus can motivate with evidence.

What we would ask its authors