Humans are characterized by an unusual level of prosociality. Despite this, considerable indirect evidence suggests that biological kinship plays an important role in altruistic behaviour. All previous reports of the influence of kin selection on human altruism have, however, used correlational (rather than experimental) designs, or imposed only a hypothetical or negligible time cost on participants. Since these research designs fail either to control for confounding variables or to meet the criteria required as a test of Hamilton's rule for kin selection (that the altruist pays a true cost), they fail to establish unequivocally whether kin selection plays a role. We show that individuals from two different cultures behave in accordance with Hamilton's rule by acting more altruistically (imposing a higher physical cost upon themselves) towards more closely related individuals. Three possible sources of confound were ruled out: generational effects, sexual attraction and reciprocity. Performance on the task however did not exhibit a perfect linear relationship with relatedness, which might reflect either the intrusion of other variables (e.g. cultural differences in the way kinship is costed) or that our behavioural measure is insufficiently sensitive to fine‐tuned differences in the way individuals view their social world. These findings provide the first unequivocal experimental evidence that kinship plays a role in moderating altruistic behaviour. Kinship thus represents a baseline against which individuals pitch other criteria (including reciprocity, prosociality, obligation and a moral sense) when deciding how to behave towards others.

Adult male Norway rats were tested in a first experiment to see whether foraging efficiency could be improved by social learning. Observers were placed in one of four conditions in which they were paired with demonstrators that either had or had not been previously trained to dig for buried carrot pieces, and in which the demonstrators either did or did not have carrot buried in the experimental enclosure. Observers in the group with trained demonstrators that did have carrot pieces buried in the experimental area during the observation period subsequently unearthed more buried carrot, did so more rapidly, and were generally more active than were the observers in the other three groups. In a second experiment, chains of transmission were established by allowing each observer to act as a demonstrator for the next naive observer. Enhanced levels of digging behavior were maintained across eight transmission episodes in three transmission groups relative to a no-transmission control group, the performance levels becoming stable after five transmission episodes at a level significantly above that of the control group. The study demonstrates that social learning and transmission mechanisms exist which might result in the diffusion of certain patterns of behavior through populations of Norway rats.The experimental study of social learning in rats began with Small's work on puzzle boxes (Small, 1899(Small, -1900. It has since been extended to many other situations, encompassing both arbitrary and artificial tasks as well as tasks that bear some relationship to the natural history of the rat (e.g., Bayroff & Lard, 1944;Galef & Wigmore, 1983;Miller & Dollard, 1941; Zajonc, 1965;Zentall & Levine, 1972). Rats are versatile feeders, eating plants, fungi, various invertebrates such as insects and molluscs, and even small vertebrates; yet they are also neophobic (Barnett, 1975;Twigg, 1975). Also, Norway rats in particular inhabit a fixed burrow system, which is used as a base from which to forage (Calhoun, 1962). The social transmission of information about foods, their palatability, location, acquisition, and processing, might partly explain the rat's success as an omnivore. This line of thought has resulted in recent years in the focus of social learning studies of the rat on feeding behaviors (Galef, 1988).Social learning has already been demonstrated to improve foraging efficiency (Galef & Wigmore, 1983). What has not been shown is whether behaviors that have been changed by social learning have the potential for becoming fixed in a rat population. For this to occur, sociallearning must be extended into repeated transmissionWe are grateful to C. Heyes of University College London. and to M. Fanselow and the anonymous referees of this journal for helpful comments. This work was carried out under a Science and Engineering postgraduate studentship to K. N. Laland. Correspondence should be addressed to H. C. Plotkin at the Department of Psychology. University College London, London WCIE 6BT, England.Copyright 1990 Psychonomic...

Backward conditioning of the nictitating membrane response of rabbits was studied. The prior exposure to unconditional stimulus-conditional stimulus (US-CS) pairings retarded subsequent forward conditioning. Control procedures eliminated latent inhibition and forward safety-signal effects as accounting for the data. The relevance of this report to other recent studies of backward conditioning is discussed.

The fundamental tenet of contemporary sociobiology, namely the assumption of a single process of evolution involving the selection of genes, is critically examined. An alternative multiple-level, multiple-process model of evolution is presented which posits that the primary process that operates via selection upon the genes cannot account for certain kinds of biological phenomena, especially complex, learned, social behaviours. The primary process has evolved subsidiary evolutionary levels and processes that act to bridge the gap between genes and these complex behaviours. The subsidiary levels are development, individual animal learning, and socioculture itself. It is argued that individual learning is pivotal to the derivation and biological analysis of culture. The differences between cultural and noncultural societies are stressed. It is concluded that such a multiple-level model of evolution can form the basis for reconciling opposing sides in the sociobiology debate.