Visualization grammar

One-line summary: High-level encoding of data to marks/channels (or JSON encodings) instead of a fixed chart-type catalog or a raw scene graph.

The insight

observable-plot and vega-lite both sit above a lower toolkit (d3 vs Vega). They make common charts short and leave bespoke scenes to the lower layer. From 2026-08-21-autoresearch-best-frontend-visualization-libraries.

From 2026-08-21-autoresearch-best-frontend-visualization-libraries: Plot “optimizes for conciseness and memorability” and uses D3 for scales, shapes, and transforms.
From 2026-08-21-autoresearch-best-frontend-visualization-libraries: Vega-Lite specs “encode data to mark properties; the compiler produces axes, legends, and scales.”
From 2026-08-21-autoresearch-best-frontend-visualization-libraries: Official Plot vs D3 tradeoff — Plot is “more efficient” and “necessarily less expressive.”
From 2026-08-21-autoresearch-best-frontend-visualization-libraries: No fetched official page ranks Plot against Vega-Lite.

Grammars are not template kits (batteries-included-chart-library) and not React primitive kits (react-visualization-layer).