Style-Shifting: Audience Design in Speech (2D)

Sociolinguistic variation is not just about which dialect a person speaks — it's about how the same speaker moves between variants from one sentence to the next. This 2D companion to the 3D style-shifting simulator keeps the same real logistic choice model, grounded in Labov's attention-to-speech work and Bell's audience-design theory, and computes a live probability of using a standard versus a vernacular word variant from three context sliders (setting formality, addressee status, social distance) and one personal trait (prestige-consciousness). Every simulated utterance is a genuine weighted random draw that travels between a speaker and listener node on a pannable, zoomable top-down map, colour-coded by variant and logged into a rolling history strip, while live readouts track the instantaneous probability and the rolling standard-usage rate. Unlike the 3D version, either node can be clicked to fire an utterance on demand, and the whole map can be freely panned and zoomed rather than orbited.