- Stump, G. T. 1990a. 'Breton inflection and the split morphology hypothesis', Hendrick (éd.), The Syntax of the Modern Celtic Languages, 97-119.
- analyse les doubles pluriels bretons comme des cas résultant de la précédence d'une opération de morphologie inflexionnelle sur une opération de morphologie dérivationnelle.
Cet article fait suite à:
- Stump, G. T. 1989a. 'A note on Breton pluralization and the Elsewhere Condition', Natural Language and Linguistic Theory, 7:261-273.
Il en existe aussi une version pour le lectorat non-anglophone.
- Stump G. 1990b. 'La morphologie bretonne et la frontière entre la flexion et la dérivation', La Bretagne Linguistique 6 :185-237. CRBC, Brest.
- Sproat, R. 1992. 'Review of The syntax of the modern Celtic languages ', Lingua 87(4), 347–370. doi:10.1016/0024-3841(92)90018-e.
"Stump’s purpose is to present Breton evidence disconfirming the 'split morphology' hypothesis, a view due especially to Anderson (1982), and more recently argued for by Perlmutter (1988) (who is also responsible for the term), wherein derivational morphology applies ‘in the lexicon’, and inflectional morphology applies ‘extra-lexically’. Inflectional morphology under this view applies after rules of syntax have set up feature structures relevant to the spelling out of inflectional morphology, whereas derivational morphology’s responsibility is merely to build words which are fed into the syntax in the base component. [...] Thus, morphology is split into a derivational and an inflectional component, with - crucially - inflectional rules applying (much) later than derivational rules. Convincing counterevidence to the strongest interpretation of such a view would come from data which showed that inflectional rules may feed derivational rules, and it is such data that Stump provides. As is de rigueur in such matters, Stump starts his discussion with the familiar set of considerations in distinguishing inflection from derivation: (14a) Inflection encodes syntactically relevant information, whereas derivation doesn’t. (14b) Inflectional processes are fully productive, derivational processes less so. (14c) Inflectional processes preclude further derivation, derivational processes do not; and so forth. [...] The problem with Stump’s position, as with most straightforward ideas on the relation between inflection and derivation, is that it would appear to be an all-or-nothing view. Once you allow inflection to feed into derivation, then one allows a whole menagerie of cases that one might generally like to avoid. Of course there are things that could be done: one might, for example try to discover what kinds of semantic considerations might make derivation from plural nouns more likely than derivation from some other inflectional category. One might also try to use semantic considerations to help explain why Breton is the marked case, something not explained in Stump’s theory. If such a venture were to succeed, then perhaps we might finally arrive at a proper understanding of such morphological issues."