In this paper, it is argued that the word-boundary is not crucial when it comes to distinguishing pragmatically motivated repetition from grammatically motivated reduplication. It is shown that both phenomena may involve entire syntactic words as their input to yield strings of identical words. Similarly, there is also evidence of repetition and reduplication operating within a given word-unit. The main difference between repetition and reduplication is tightly connected to the distinction of pragmatics and grammar in the sense that repetition fullfills exclusively prgamatic tasks whereas reduplication is responsible for the expression of grammatical categories. Empirical evidence of this distinction in genetically, areally, and typologically different languages is discussed. The approach employs a qualitative methodology in order to evaluate the data synchronically.