This paper discusses spontaneous communication and its implications for understanding empathy and altruism. The question of the possibility of "true" altruism-giving up one's genetic potential in favor of the genetic potential of another-is a fundamental issue common to the biological, behavioral, and social sciences. Darwin regarded "social instincts and sympathies" to be critical to the social order, but the possibility of biologically-based prosocial motives and emotions was questioned when selection was interpreted as operating at the level of the gene. In the selfish gene hypothesis, Dawkins argued that the unit of evolutionary selection must be an active, germ-line replicator: a unit whose activities determine whether copies of it are made across evolutionary timescales. He argued that the only active replicator existing across evolutionary timescales is the gene, so that the "selfish gene" is a replicator motivated only to make copies of itself. The communicative gene hypothesis notes that genes function by communicating, and the phenotype communication involves not only the individual sending and receiving abilities of the individual genes involved, but also the relationship between them relative to other genes. Therefore the selection of communication as phenotype involves the selection of individual genes and also their relationship. Relationships become replicators, and are selected across evolutionary timescales including social relationships (e.g., sex, nurturance, dominance-submission). An interesting implication of this view: apparent altruism has been interpreted by selfish gene theorists as due to kin selection and reciprocity, in which the survival of kin and comrade indirectly favor the genetic potential of the altruist. From the viewpoint of the communicative gene hypothesis, rather than underlying altruism, kin selection and reciprocity are ways of restricting altruism to kin and comrade: they are mechanisms not of altruism but of xenophobia.