EN
This paper is a rejoinder to István Pete's two articles in previous numbers of this periodical, in which he intends to 'redefine' the concept of the morpheme, and within it that of the zero morpheme. I take issue with him on several counts; primarily by arguing against his view that the morpheme can be defined as a theory-independent notion, and claim that, just as in the case of the phoneme and most other terms in linguistics, what we understand by morpheme depends on the theses, principles, etc., of particular theories. In a modular grammar, for example, it is the needs of the Lexicon, i.e., the idiosyncratic elements to be listed, the 'listemes' of di Sciullo and Williams (1987), that determine the basic units, which then other approaches might call morphemes. The rest is a defense of the analyses put forward in Kenesei (2000) and criticised by Pete (2004a, 2004b).