Fungal Intelligence

Can a fungus think? The question sounds absurd — fungi have no neurons, no brain, no nervous system of any kind. But research over the past two decades has documented behaviors in mycelial networks that are difficult to describe without cognitive vocabulary: problem-solving, memory, decision-making, anticipation. The study of fungal intelligence asks what these behaviors mean and what they reveal about the relationship between cognition and network structure.

The most dramatic demonstration came from Toshiyuki Nakagaki’s work with the slime mold Physarum polycephalum (technically a protist, not a fungus, but one that forms mycelium-like networks). When placed in a maze with food at two exits, Physarum extended its network throughout the maze, then pruned back all branches except the shortest path between the food sources. When food sources were arranged to match the positions of cities around Tokyo, the organism produced a network strikingly similar to the actual Tokyo rail system — an optimization that human engineers achieved only with decades of planning. Similar network optimization behaviors have been observed in true fungi. Phanerochaete velutina and other wood-decay Basidiomycota form foraging networks that balance exploration (extending into new territory) against exploitation (reinforcing connections to discovered resources), reallocating cytoplasm from unproductive branches to productive ones.

These behaviors emerge without central coordination. No part of the mycelium directs the whole. Instead, local interactions between hyphae — chemical signaling, cytoplasmic streaming, differential growth and retraction — produce global patterns that are adaptive at the network level. When a nutrient source is discovered, hyphae near it thicken and increase transport capacity. Hyphae extending into empty territory thin and may be withdrawn. The network reorganizes itself continuously, and the result is a structure that looks like it was designed to solve a transport problem. This is distributed cognition: intelligent behavior arising from the interactions of individually simple components, without a central processor.

The memory question is particularly striking. Physarum can learn. Audrey Dussutour and colleagues showed that the organism can habituate to aversive but harmless stimuli (quinine, caffeine), learning to cross chemical barriers it initially avoided. This learned tolerance persists and can even be transferred to naive individuals through cell fusion — a form of somatic memory transmission that bypasses any genetic mechanism. Whether true fungi exhibit comparable memory remains an open question, but mycelial networks do show path-dependent behavior: a network that has previously colonized a substrate retains structural biases that influence future growth patterns. The network’s history is encoded in its architecture.

For the vault’s relational framework, fungal intelligence is significant because it demonstrates that cognitive-like behavior does not require a subject — a bounded individual with internal representations. The mycelial network has no inside where representations could be stored, no executive that could consult them. Its “intelligence” is relational: it consists entirely in the pattern of connections between hyphae, the flows of material between them, and the way these patterns change in response to environmental signs. This connects to autopoiesis — the network continuously produces and reorganizes itself — and to biosemiotics — the network’s behavior is sign-mediated, driven by chemical detection and response. The fungal umwelt is navigated not by a perceiving subject but by a distributed network whose perceptual and effector capacities are one and the same: growth toward, or retraction from, chemical signals.

Related

• Mycelial Networks — the physical substrate of distributed fungal cognition
• Fungal Chemical Ecology — the chemical semiosis that mediates network behavior
• Anastomosis — hyphal fusion, which creates the network connectivity that enables distributed processing
• Umwelt — the organism’s sign-world, radically distributed in fungi
• Autopoiesis — the self-producing character of networks that reorganize themselves
• Niche Construction — intelligent foraging as a form of environmental modification
• Morphogenesis — network form emerging from local interactions without central plan