Thread by @jeffrey_heer

For a decade+, grammar-of-graphics approaches (ggplot, Tableau, #d3js, Vega/Altair) have been a leading way to make visualizations. Beyond chart templates & low-level programming, are there compelling alternatives? Or does the future lie in abstractions on top of these grammars?

There's exciting research work on new/extended grammars, including:
- probability expressions 📊 (mjskay.com/papers/chi2020…)
- responsive charts 📲 (

https://twitter.com/hyeok__kim/status/1506395769615568904

)
- animation 🕺 (idl.cs.washington.edu/papers/gemini2, gganimate, and new @vega_vis work coming to VIS'22 from @mitvis)

.@_mcnutt_ has written a valuable survey of JSON-style grammars (arxiv.org/pdf/2207.07998…), concluding there is "No Grammar to Rule Them All". Should we expect a proliferating multiverse of visualization grammar variants? Where might we look beyond (or building on) such grammars?

I'm increasingly interested in higher-level support for analysis tasks. Grammars are a powerful way for (both people and machines!) to say "make this chart". But most grammars are agnostic to the larger task and intent. Could we (should we?) bridge that gap?

(And for any friends wondering: Yes, I should be working on #chi2023 papers right now. And no, I'm not writing a CHI paper on this topic... #procrastination 😅)

p.s. thinking about this also reminds me how much I miss Lee Wilkinson (RIP). He was always bubbling with ideas on this topic... before he passed we had been discussing multimodal grammars, including sonification, tactile displays, etc. Important to also think beyond the visual.