The relation between word-formation and syntax and whether they form distinct domains of grammar or not has been discussed controversially in different theoretical frameworks. The answer to this question is closely connected to the languages under discussion, among
other things, because languages seem to differ considerably in this regard. The discussion in this paper focuses on nominal compounds and phrases. On the basis of a great variety of data from a total of 14 European languages, it is argued that the relation between compounds and phrases, and,
more generally, between word formation and syntax, should be characterized not in terms of a categorical but instead in terms of a gradient distinction.