In 1994, Fuqua, Winans, and Greenberg named a widespread bacterial behavior quorum sensing: cells release autoinducer molecules and switch on cooperative behaviors when concentration crosses a threshold — as if waiting for a vote to reach quorum.
In 2002, Rosemary Redfield reanalyzed the mechanism. Autoinducer concentration doesn't measure cell density. It measures density relative to local diffusion rate. A cell in a crevice (restricted diffusion) accumulates autoinducer even at low density. A cell in open water disperses its signal even at high density. The bacteria aren't sensing how many neighbors they have — they're sensing whether diffusion is restricted, which is a statement about local geometry.
Both panels below use identical cell counts and emission rates. The difference is diffusion geometry: autoinducer disperses rapidly on the left (open water), and accumulates on the right (restricted environment). Watch the right panel activate over ~3 seconds.
Adjusting threshold changes when cells activate. At low thresholds, even the open panel may show some activity near dense clusters. At high thresholds, even the closed panel may not fully activate — modeling environments where the cooperative behavior requires a very strong signal.
Redfield's 2002 paper argued that "quorum sensing" is a misnomer for most environments. What bacteria actually monitor is whether they are in a setting where secreted molecules can accumulate — which tells them whether secreting extracellular enzymes or signaling molecules will be locally effective, not whether a community has assembled.
Hense et al. (2007) refined this further as efficiency sensing: the autoinducer concentration encodes a ratio of cell density to mass transfer rate. The cell isn't asking "are there enough of us?" — it's asking "will my secretions stay local long enough to matter?"
→ entry-587: The Quorum · quorum sensing simulation · chemotaxis