Suppletion is a phenomenon which is unique to morphology. It represents extreme non-iconicity in the relation between form and function. Here we deal with what we call proper suppletion, i.e. where one paradigm has resulted from the interaction of two previous paradigms. Our aim is to establish what the conditions are that give rise to this (partial) merger of paradigms. We provide a detailed account of the development of an adjective in the Mainland Scandinavian languages which unexpectedly shows suppletion with respect to number (and in some cases definiteness). We conclude that development of suppletion is driven by semantic properties, and that frequency is in fact a derived property. Furthermore we argue that there must be an asymmetry between the participants, with one element being dominant.
It has been argued that the development of the Swedish possessive -s constitutes an example of degrammaticalisation because it has changed from being an affix to being a clitic (Norde 1997; 2001a, b). I argue that a simple distinction between affix and clitic cannot capture the development of this element, instead a distinction in two dimensions need to be made; PLACEMENT needs to be distinguished from DEGREE OF ATTACHMENT. Furthermore, I claim that the distinction between agreement and once only marking represents yet another dimension. With respect to the Swedish possessive, as Norde (1997) shows, there has been a clear change from agreement to once only marking; however, I argue that this change does not tell us anything about the element's morphological status. With respect to placement, -s is still subject to competing constraints and with respect to attachment, there is some evidence of morphological interaction between the -s and the word to which it attaches. It is then inappropriate to call the Swedish possessive -s a clitic and to refer to the change which it has undergone as degrammaticalisation.
In this paper, we argue for a severely restricted use of zero morphemes and against the use of features which are not strictly motivated by the distribution of the element with which the feature is assumed to be associated. We consider an element to have a particular value for a feature only if the element in question can occur exclusively in environments where that feature value is present on some other element and never in environments where that feature has another value. We apply this definition of features and feature values to noun phrases in Germanic, particularly to Dutch and Danish. We use Optimality Theory in conjunction with our approach to features and values and show that in this way, the distribution of features relating to definiteness, number and gender can be correctly predicted for these languages. We then consider the consequences this view of features has for the choice of an approach to morphology.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.