Between the ages of eighteen and thirty months, various markers – often called “incorrect forms” in the literature - are used in subject position by children to designate themselves: null-form (want), preverbal vowels (a want), name (Peter do), he/she (he goed to sleep), me (me do this), my (my go home), you (you mixed the egg!). Between thirty and thirty-six months, the use of I is stabilised and the other markers tend to disappear in subject position. At the same time, the children start manipulating various tenses, moods and aspects. How can the use of these competing forms be explained? Are there mandatory stages that prepare for the use of "I"? What are the discursive, psychological, semantic, pragmatic, linguistic problems children have to face in order to catch and keep this tiny, slippery pronoun? In this article, we shall analyse the data of Peter, an English speaking little boy, between the ages of 18 and 36 months. We will try to show how creative, his “incorrect”, or rather his “non-adult” uses of self-designations can be and how they help him master the semantic and syntactic complexity of the pronoun I (since it identifies the speaker, the grammatical subject, and the actual little boy named Peter). After the child has built a very organised system in which each self-word has its function, he will finally join these various functions in one marker, I. At the end of the acquisition process, the child finally masters the adult system, and shows a full appropriation of language and an equal sharing in co-enunciation.