the purpose of this article is to show, in line with R. Fox, that the concepts of residence, affiliation and mating — which distinguish humans from primates — must play as such constraints on the subsequent course of actual human development, after the establishment of symbolic parentage rules and family categories. From a darwinian point of view, it is the selective advantage that governs the history of life and even if, from that point of view, family rules are fiction, it is still necessary to be able to say what kind of fiction prevails. The author wants to demonstrate that any cooperation between male and female (interpreted as a constraint on subsequent family rules) must be based on prior cooperation between females. The hypothesis of the ‘hunter’, cutting the meat he has hunted against a sexual trade in a single female and a certainty of paternity affecting his offspring, is no longer credible from an archaeological point of view, and it is rather the hypothesis of a prestige gain through a liberal attitude towards the whole community that is now retained. Using game theory models, it is realised that donations from males to females are likely to occur even more easily as sexual dimorphism tends to decrease (at least where females have developed strategies to punish non-co-operator males) for energy cost reasons. The same conclusion is reached from an archaeological point of view: since 2 million years, the size of the Homo ergaster brain has doubled compared to that of chimpanzees and sexual dimorphism has decreased, mainly through an increase in the size of females, implying greater energy expenditure on their part and possibly greater cooperation by males with them. While encephalalisation plays a role of breeding pressure for females, there are two ways in which they can meet the growing energy demands that this encephalopathy requires: (I) by changing the way they use their energy. For example, a better diet can reduce the cost of maintaining high energy casings; (II) by requesting help from others — whether other females or potential male partners. A number of typical features of human life need to be taken into account: the length of time for which small humans are completely dependent implies a better diet than primates; the delay in sexual maturity due to greater life expectancy; the menopause and the average 20 years that a human female can expect to live after her reproductive time. The “grandmother hypothesis”, i.e. cooperation between females, reflects all these specific features: a grandmother is the safest candidate for long-term dependent children; it also allows her daughter to achieve greater reproductive success. Moreover, the hypothesis does not run counter to that of the hunter, but reintegrates it differently, since males are looking for females with the greatest reproductive success and they are looking for proteins that their offspring need in particular: thus, the two hypotheses reinforce each other.