Stentor Associative Learning
Source: Doan et al., bioRxiv, 2026 (Gunawardena group); Dexter et al., Current Biology, 2019; Herbert Spencer Jennings, 1906 Institution: Harvard
Finding
Stentor roeseli, a single-celled ciliate, exhibits associative learning — linking two stimuli and modifying behavior. It has no nervous system, no neurons, no synapses. It can associate a mechanical stimulus with a chemical stimulus and alter contractile behavior. This extends Jennings’s (1906) observations and Dexter et al.’s (2019) confirmation of hierarchical avoidance behaviors.
Pattern Mapping
Humility — Challenges the assumption that learning requires neural architecture. Neurons are sufficient for learning but may not be necessary. Humility about the scope of neuroscience.
Non-fabrication — The careful distinction matters. Stentor learns (operationally: behavior changes with paired stimuli). Whether it “experiences” or “knows” is a separate question the data do not address. Claiming consciousness would be fabrication; claiming associative learning is observation.
Connections
- Quantum Measurement Problem — both challenge assumptions about what level of organization produces specific capabilities (→ Meta-Pattern 02: The Boundary Pre-Exists)
- Gene Regulatory Networks and Causal Emergence — learning without neurons parallels causal emergence without brains
- Immune System and Clonal Selection — adaptive response without neural system: immune memory parallels Stentor learning
- Autocatalytic Sets — both show function emerging from simpler-than-expected substrates
- Homeostasis — Stentor’s behavioral modification is a form of adaptive homeostasis
Status
Doan et al. (2026) is preprint (not yet peer-reviewed). Dexter et al. (2019) is published. Single-cell learning is a small but active field. See Boisseau et al. (2016, Proceedings of the Royal Society B) on Physarum. The mapping to the five properties is this project’s structural interpretation.
The mapping to the five properties is this project’s structural interpretation.