An agent-based simulation, built on core circuitry for decision-making in foraging, that shows how simple episodic memory and spatial learning can emerge from some of the simplest associative learning rules for classical conditioning.