Whether we twitter, dance or send out scents, most creatures communicate with a suite of signals only appreciated by their own species. Many creatures have co-opted sound, colour and movement for communication, but mormyrid fish have opted for a relatively unconventional communication mode; they emit weak electric pulses. Bruce Carlson explains that each fish emits its own personalised electric bleep, but when Carlson recorded the fish's electric conversations, he realised that the fish also bleep in one of four distinct calling patterns. When resting, the fish discharge a simple train of low frequency pulses, but when they burst into action, they have the choice of three electric call patterns: scallops, accelerations and rasps. Knowing that the fish's conversational repertoire was limited to four `calls', Carlson and his advisor Carl Hopkins decided to discover how the fish regulates its electric voice, and found that mormyrids use a single negative feedback loop to control the entire signalling system(p. 1073).But before he could work out how the circuitry controls the fish's call patterns, Carlson had to trace the circuit's intricate network of neurons. From previous work, it was already known that a region of the brain, known as the command nucleus, directly controls each electrical discharge; there's a 1:1 correlation between an electronic discharge and command nucleus activity. But when he mapped the regions of the fish's brain that triggered each electric organ discharge Carlson realised that the command nucleus was in turn regulated by both the midbrain precommand nucleus (PCN) and the dorsal posterior nucleus (DP). Carlson was puzzled; why was the command nucleus regulated by two upstream command centres?Stimulating the PCN with the neurotransmitter glutamate, Carlson recorded the resulting electric discharge, and realised that the PCN was driving call patterns similar to the scallop signal. Repeating the experiments but stimulating the DP, Carlson recorded signals similar to the fish's acceleration call. Each centre controls a specific discharge signal, and they probably work together to produce the third.Carlson's neuron mapping had also thrown up an additional neural connection between the command nucleus and the PCN and DP; a feedback loop, that went via the ventroposterior nucleus of the torus semicircularis (VP). Carlson wondered why the fish might use a feedback loop from the command nucleus to regulate both driving nuclei. Was it inhibiting the signals from the PCN and DP to maintain the fish's gently repeating resting calls? If the feedback loop was inhibitory, cutting it off should provoke the fish to being uncontrollably discharging. Carlson decided to pharmacologically cut off the fish's feedback loop, and see what signals they began sending.Blocking the feedback signal from the VP, Carlson measured the electric organ discharge. Sure enough, `the fish suddenly started bursting like crazy'says Carlson, `there was a very, very dramatic shift of activity... it was clear that negative feedback plays a big role' he added. And when he took a closer look at the way the feedback signal interacted with both the DP and PCN, he realised that the signal also sets up differences between the two driving nuclei, resulting in the different calling patterns. `The negative feedback sets up the entire system' says Carlson `keeping it regular, but also allows bursts of electrical discharge'.