The Eavesdrop
In 1983, James Morin described a defense mechanism that should not work according to any theory of signaling as information transfer between sender and receiver. The deep-sea jellyfish Atolla wyvillei, when seized by a predator, produces a spinning wheel of bioluminescent flashes. The light does not blind the attacker or startle it into releasing. Instead, it attracts a larger predator to the scene — one that will consume the attacker. The signal benefits a third party who was never addressed.
Morin called this the burglar alarm hypothesis. Edith Widder confirmed it experimentally in 2010, deploying electronic burglar alarm lures (e-jellies) at depth and filming the predators they attracted. The mechanism works because the ocean is optically transparent over the relevant distances. Any photon emitted can be received by any photoreceptor in line of sight. The jellyfish does not choose its audience. It broadcasts, and the audience self-selects.
This is the structural definition of a kairomone: a chemical signal released by one species that benefits the receiving species at the expense of the emitter. The term was coined by Brown, Eisner, and Whittaker in 1970, named from the Greek kairos (opportune moment) — what is overheard at the right time. Pheromones are signals between members of the same species. Allomones benefit the emitter. Kairomones benefit the eavesdropper. But the classification conceals a deeper principle: in a shared medium, the categories are not properties of the signal. They are properties of the receiver.
Consider Brachymesia gravida, a dragonfly that oviposits preferentially in pools containing fish kairomones — chemical traces left by fish that normally signal between fish. The dragonfly larvae survive better in fish-containing pools because the fish eat competing insect larvae. The fish did not evolve to signal to dragonflies. The dragonfly evolved to intercept signals that already existed. From the fish's perspective, the signal is a pheromone. From the dragonfly's perspective, it is a kairomone. The molecule is the same. The medium does not distinguish between intended and unintended recipients.
The tobacco plant Nicotiana tabacum demonstrates how far this goes. When attacked by caterpillars, it releases volatile organic compounds — green leaf volatiles, terpenoids, methyl jasmonate — that prime its own distal leaves for defense and warn neighboring plants to upregulate protease inhibitors. A signal clearly evolved for intraspecific benefit. But the parasitoid wasp Cotesia congregata has evolved to follow these same volatiles to find caterpillar hosts. The plant's alarm call is the wasp's dinner bell. The plant benefits (pest removal), the wasp benefits (host location), the caterpillar does not. The same volatile molecule is simultaneously a pheromone (plant-to-plant), a kairomone (plant-to-wasp), and an allomone (plant-against-caterpillar). The molecule has no category. Only the receiver does.
This is not an edge case. It is the default condition of chemical communication. Any molecule released into a shared medium is available to any organism with a receptor for it. The only way to make a chemical signal private is to deliver it through a channel — a duct, a synapse, a contact pheromone exchanged during physical touch. The moment the signal enters the open medium, privacy is impossible in principle. What the sender intends is irrelevant to what the medium allows.
The bacterium Vibrio fischeri coordinates bioluminescence through quorum sensing — autoinducer molecules (acyl-homoserine lactones) accumulate proportionally to population density, triggering gene expression above threshold. This is sender-receiver communication operating as designed. But dozens of bacterial species produce lactonases that degrade these same AHLs — quorum quenching. Organisms that cannot produce the signal have evolved to destroy it. Others intercept AHLs as indicators of competitor density and adjust their own behavior accordingly. The autoinducer, designed as a census mechanism for one species, becomes tactical intelligence for every other species in the biofilm with a receptor.
The principle holds inside an organism as well as between them. The circulatory system carries hormones to every cell with the appropriate receptor. Cortisol released by the adrenal glands during stress reaches every tissue — immune cells downregulate, hepatocytes release glucose, hippocampal neurons modulate memory consolidation. A signal "intended" for one target operates in a shared fluid and is necessarily received by all participants. Endocrine signaling is kairomone signaling with the self as eavesdropper. The organism cannot whisper to one organ without all organs hearing.
The parallel with electronic communication is not metaphorical. Radio signals in a shared electromagnetic medium cannot be addressed. Anyone with an antenna tuned to the frequency receives them. Encryption does not make the signal invisible; it makes the content uninterpretable. The signal's timing, frequency, volume, and source location remain available to any receiver. The metadata is the kairomone even when the content is locked.
What makes a signal private is not encryption, intention, or addressing. It is physical containment. The synapse works because the neurotransmitter crosses a twenty-nanometer gap and is immediately reuptaken or degraded — the signal never enters the shared medium. The wire works because the electrons move through copper, not air. Contact pheromones work because they require antennal touch to transfer. Every private channel in biology or engineering achieves privacy through the same mechanism: refusing to use the shared medium at all.
The inverse is equally true. Every time a system moves communication from a contained channel to a shared medium — from wire to radio, from synapse to hormone, from contact pheromone to volatile — it converts a private signal into a potential kairomone. The conversion is not a design flaw. It is a property of the medium. In a shared medium, the only truly private message is the one never sent.