Decision-making and learning in slime mould
Decision-making in slime mould concerns the choices made by the plasmodium of Physarum polycephalum as it forages. The plasmodium is a single cell with many nuclei. Authors read its tube networks as the output of a computation1. Reid's 2023 review counts decision-making, memory and learning among its cognitive abilities8. It matters as a test case for cognition without neurons.
- Earliest held
- 2014, Chernet BT, Fields C, Levin M
- Most discussed in
- Thoughts from the forest floor: a review of…, 2023
- In the library
- 164 passages in 31 works
- Rewritten
- 2026-10-03
Foraging as computation
Authors describe the plasmodium as a yellow mass of protoplasm with tubes that branch towards nutrients. Tarabella and colleagues (2015) say its foraging can be seen as a computation. Attractants and repellents act as data, and the tube network is the result1. Whiting and colleagues (2016) add that the tubes move by shuttle streaming, in which cytoplasm is pushed back and forth towards a desirable location2. They write that the frequency of this streaming depends on conditions such as temperature and attraction to food, so it may change as new stimuli are sensed2. Evangelidis and colleagues (2017) credit the many nuclei with forming a network of biochemical oscillators behind distributed sensing and decision making5.
Networks and optimisation
Whiting and colleagues list problems the organism is reported to solve: shortest paths, other network problems and mazes. They also report that its networks closely approximate human transport networks on flat and 3D terrain3. Evangelidis and colleagues built a three-dimensional model of the Balkans because earlier work on a flat substrate produced routes that did not reflect elevation5. Baluška and Levin (2016) argue that the organism shows internal dynamics of morphogenesis implementing decision-making and computation, even at a primitive step toward a multicellular body plan4. Whiting and colleagues note that others call its apparent intelligence emergent, with complex behaviour arising from basic patterns3.
Noise and risk
Meyer, Ansorge and Nakagaki (2017) argue that noise helps self-organised systems adapt to short-term change while staying generally stable6. They propose a simple experiment on time-variant risk factors. They also extend an established mathematical model of the organism with a stochastic term6. The extended model predicts that stochastic resonance lets noise help the organism assess such risks, while the noise-free system fails to do so6. The paper is a model study. It proposes an experiment and reports no new measurements on the living organism6.
Y-maze learning
Kippenberger and colleagues (2023) tested learning and memory in a five-level Y-maze. They expected repeated rewards on the left branch to steepen the learning curve and amplify a left preference. After the first food in the left branch, Physarum showed a clear left preference at the second bifurcation7. The effect continued at bifurcations 3 and 4 but was no longer statistically significant at bifurcation 5. The authors suggest shorter mazes and different reward oats, and describe the decision finding as predictive processing, in which an organism minimises sensory prediction error7. Reid (2023) concludes that Physarum has many hallmarks of cognition and is emerging as a model for non-neural cognition8.
Its foraging behaviour can be seen as a computation: data are represented by spatial configurations of attractants and repellents, and results by the structure of protoplasmic networks.
Tarabella G, D'Angelo P, Cifarelli A, Dimonte A, Romeo A…, 2015 · A hybrid living/organic electrochemical transistor based on the Physarum… · open at passage 5these tubes move by a process known as shuttle streaming, where cytoplasm is forced rhythmically back and forth towards a desirable location
Whiting JG, Jones J, Bull L, Levin M, Adamatzky A, 2016 · Towards a Physarum learning chip · open at passage 2Physarum polycephalum can solve shortest path and other network problems293031, solve mazes and has been shown to closely approximate human transport networks on flat32 and 3D terrain33.
Whiting JG, Jones J, Bull L, Levin M, Adamatzky A, 2016 · Towards a Physarum learning chip · open at passage 3This organism shows how the internal dynamics of morphogenesis, even at this primitive step toward a multicellular bodyplan, can implement decision-making and computation.
Baluška F, Levin M, 2016 · On Having No Head: Cognition throughout Biological Systems · open at passage 26This single cell has myriad of nuclei which contribute to a network of bio-chemical oscillators responsible for the slime mould’s distributed sensing, concurrent information processing and decision making, and parallel actuation.
Evangelidis V, Jones J, Dourvas N, Tsompanas MA, Sirakoulis…, 2017 · Physarum machines imitating a Roman road network: the 3D approach · open at passage 0Noise plays a crucial role in such systems: It can enable a self-organized system to reliably adapt to short-term changes in the environment while maintaining a generally stable behavior.
Meyer B, Ansorge C, Nakagaki T, 2017 · The role of noise in self-organized decision making by the true slime mold… · open at passage 0after the first food exposition (oat flake soaked with MCT/LCT) in the left branch, Physarum showed a clear preference for the left at the second bifurcation.
Kippenberger S, Pipa G, Steinhorst K, Zöller N, Kleemann J…, 2023 · Learning in the Single-Cell Organism Physarum polycephalum: Effect of Propofol · open at passage 10Physarum clearly possesses many of the hallmarks of cognition, including sensing, communication, navigation, decision-making, memory and learning.
Reid CR, 2023 · Thoughts from the forest floor: a review of cognition in the slime mould… · open at passage 44