The various different occurrences of -er in the morphology of German (e.g. in agentive and instrumental nouns, in comparatives, iteratives and intensifiers) are generally treated as being homonymous. In other words, just as in the case of full lexemes (e.g. Kiefer(knochen) 'jaw bone' vs. Kiefer(baum) 'fir tree'), it is assumed that different suffixes exist and that it is merely coincidental that they have an identical phonetic realisation. As far back as Germanic, the source of this identity-of-form has been taken to lie in reduction of the final syllable, whereby the process of phonological reduction has created identical suffixes from ones that were previously formally distinct. Our aim is to illustrate that in the case of nominal -er forms, there is good reason not to attribute identity of form simply to the 'blind application' of phonological rules. On the one hand, there are various aspects of the historical development which simply cannot be accounted for by recourse to phonological rules. This will be made clear through analysis of historical grammars and related literature. On the other hand, those categories that are formed with -er share common grammatico-semantic features. This too is often hinted at in the historical grammars and may be corroborated by an analysis of the grammatical features involved. Our findings thus motivate the following claim: The emergence of identical forms is not coincidental. Rather, alongside phonological rules, the emergence of identical forms is driven by more global principles of structure-formation, in particular by the principle of 'one form = one function'. Instead of different -er suffixes, there is in fact just one suffix which symbolises the iterative feature structure common to the -er marked categories.