Imitation has been demonstrated in apes, monkeys, and birds. In accordance with some recent theories (e.g. Associative Learning (ASL) Theory), imitative capacity may depend on cognitive mechanisms (learned perceptual-motor links), which are likely to be available in various non-human species. Human imitation, however, shows some intriguing characteristics: It is not about automatically copying some others´ behavior, but shows context-dependency in at least two ways: 1) It seems to involve some selective, inferential processes, taking not only the performed behavior but also its outcomes and constraints into account. 2) It seems to be interlinked with and affected by coordinated gazing behavior and other social cues of the demonstrator and the learner. While non-human research has tried to determine whether and which species of animals are able to imitate others´ behavior, human research focuses on the question which specialities of human imitation and which other characteristics of human social information transmission might have contributed to the evolution of human culture. Candidates are the capability for reasoning, which goes beyond associative psychological processes, and the attentional coordination of companions, which makes social information transmission more effective. Both marmosets and dogs have been found to engage in different forms of social learning, but the above aspects have hardly been examined, although they may provide valuable non-human models with the special interest of the distinctive evolutionary histories of these two species.