Providing information can be a stable non-cooperative evolutionary strategy

Human language is still an embarrassment for evolutionary theory, as the speaker’s benefit remains unclear. The willingness to communicate information is shown here to be an evolutionary stable strategy (ESS), even if acquiring original information from the environment involves significant cost and communicating it provides no material benefit to addressees. In this study, communication is used to advertise the emitter’s ability to obtain novel information. We found that communication strategies can take two forms, competitive and uniform, that these two strategies are stable and that they necessarily coexist.

💡 Research Summary

The paper tackles the long‑standing puzzle of why humans invest heavily in spoken language even though the speaker receives no obvious material benefit. The author argues that providing information can be an Evolutionarily Stable Strategy (ESS) when the act of communicating serves as a costly signal advertising the speaker’s ability to acquire novel information. Two distinct communication strategies emerge from a formal model: a competitive strategy, in which signal intensity scales continuously with an individual’s “quality” (information‑gathering competence), and a uniform strategy, in which high‑quality individuals converge on a single, fixed signal level. Both strategies are shown to be evolutionarily stable and to coexist.

The paper begins by reviewing nine empirical observations about human language (e.g., 30 % of waking time spent talking, low material payoff for listeners, broadcast‑like nature of speech, early emergence in ontogeny, lack of sex differences, and the uniqueness of language among primates). Existing altruism theories—group selection, parochial altruism, reciprocal cooperation, indirect reciprocity, kin selection, sexual selection, and classic costly signaling—fail to account for one or more of these observations, especially the ubiquity of speech (O1) and its apparent lack of direct benefit to listeners (O3).

To overcome these shortcomings, the author adapts costly signaling theory (CST) into a “Friendship Signaling Model” (FSM). The model adds two key hypotheses: (1) individuals evaluate each other’s signaling performance symmetrically before forming a friendship, and (2) the amount of time an individual can allocate to a friend (the “social offer” r) limits the strength of the bond. When two agents A and B meet, A offers a product of its signal s(q_A) and the time share r_i to B; B does the same. This bilateral negotiation creates a “social relationship market” where partners are matched based on the combined signal‑time offer.

In the competitive ESS, the benefit to an individual of quality q is B_c(q)=P(q)−C·s(q)/q, where P(q) is the profit obtained from an alliance with a partner of similar competence, and C is the cost coefficient. Setting the derivative of B_c to zero yields the equilibrium condition s′(q)=q·P′(q)/C, which integrates to s(q)=