One very basic dichotomy in language is that between presupposed and non-presupposed information. Someone who utters the sentence It stopped raining implies that it was raining before, but this is not asserted: the speakers choice of words indicates that h e is taking it for granted that it was raining before; his statement is that it stopped. Languages offer a variety of means for marking the distinction between presupposed and non-presupposed content. Examples are transition verbs (stop, begin) and focus particles (too, even). These devices play a key role in communication, for they serve to signal the division between backgrounded information and the main point of an utterance. In linguistics and philosophy, presuppositions have been studied for decades, but experimental psychologists seem to have overlooked the phenomenon, and nothing is known about how presuppositions are processed.The proposed project would be the first to address the question how and when children learn to deal with presuppositions. The project consists of three studies. The first is a corpus study whose aim is to investigate when and how children learn to use presupposition-inducing expressions. The second is an experimental study designed to determine when children become attuned to t he fact that presuppositions exhibit so-called projection behaviour, which is widely agreed to be their distinctive property. The third study uses a lexical-learning paradigm to assess if young children (2-3 and 4-5 years) can spontaneously distinguish between presupposed and non-presupposed information.The project is to be carried out at the Max Planck Institute for Evolutionary Anthropology (Leipzig), which ranks among the best environments worldwide for research in language acquisition, especially with the proposed methodologies. The project will provide the applicant with a unique training opportunity in corpus research and experimental research with very young children.