Communication requires a signaler, a signal, and a receiver.   The receiver responds (sometimes) to the signal produced by the signaler.

Communication is honest (signals are honest) when both signaler and receiver benefit.   When discussing the evolution of communication, benefits and costs are measured in terms of the spread of alleles in populations.   For instance, receivers benefit from communication when alleles associated with responding to signals spread faster than those associated with not responding.   Signalers benefit in an analogous way.

Communication is dishonest (signals are dishonest) when only the signaler (or only the reciever) benefits and its partner incurs a cost.   Dishonest communication might occur when a signaler tricks a receiver into responding in a way that is against its best interests but in a way that benefits the signaler (for example, a suboptimal male tricks a female into mating with him).   Conversely, a receiver might eavesdrop on signals intended for others, in a way that benefits the eavesdropper but not the signaler (for example, a satellite male approaches an advertising male and then intercepts females arriving to mate with the signaler).

Honesty in communication can refer either to a signaling system (categories of signalers and receivers and defined signals) or to a particular instance of signaling (a particular signaler, receiver, and signal at a particular time and place).   A signaling system might be honest on average and yet include instances of dishonesty.

A widely read paper by Dawkins and Krebs (1978) maintained that communication consists mostly of individuals manipulating each other for their own benefits rather than exchanging information for their mutual benefit.   In other words, most communication was dishonest.   Furthermore, dishonesty is what you would expect to evolve as a general rule.

Zahavi about the same time (1975, 1977) proposed that communication is honest, but only when signals have costs, often extravagant costs, which he called handicaps.   Cost demonstrated that the signaler was serious.   For example, only a male in good condition can signal in expensive ways and still survive as well as a male in poor condition.

After a decade of mathematical analysis, most of which suggested that Zahavi's "Handicap Principle" did not work (costly signals did not evolve), several papers in the early 1990's claimed the opposite.   The most influential paper was by Grafen (1990), which attempted to prove a mathematical theorem that honest signaling occurred if and only if signals were costly.   This effort proved to have a mathematical flaw.

Another attempt to demonstrate that honest signaling occurred only when signals were costly used an appealing story of human altruism -- Sir Philip Sidney's offer of his water bottle to a dying soldier on a (mostly forgotten) battlefield.   This paper by John Maynard Smith (family name Maynard Smith) was widely accepted as a clear demonstration that Zahavi's Handicap Principle was correct -- honest signaling occurs only if signals are costly. [SEE THE NEXT FILE]