Disjunctive ordering and French morphology

Conclusion

A principle such as disjunctive ordering definitely has a considerable appeal and appears to play an important role in many descriptions. While its role in morphology has been seriously questioned by Janda and Sandoval (1984), Anderson (1986) shows conclusively that the cases that pose serious potential problems can be given alternative accounts which resolve the apparent contradiction. The case studies presented here, however, show that not all problems have been solved satisfactorily.

In one case — that of Picard SUBJ — a modular version of blocking appears to be more adequate. Modular blocking, however, lacks the generality which makes blocking so appealing as an explanatory principle. Furthermore, the consequences of the modularity have not yet been thoroughly investigated. Will the modularity not exclude some of the cases where a general version of blocking seemed appropriate? If modular blocking correctly allows the SUBJ marker /∫/ in Picard to be added to a SUBJ stem /swε/ of être ‘to be’, how will one explain that the plural marker -s in English cannot be added to plural stems such as feet or people, which Kiparsky, for instance, uses as clear instances of blocking?

One question which should be more thoroughly investigated concerns the status of blocking. It is really a strict principle which regulates the internal organization of grammar or rather a markedness condition — a default case — which can always be overruled through explicit specification, as suggested by Kiparksy in his discussion of doublets such as croci/crocuses in English? If one accepts the latter interpretation, then one must make precise what can constitute an empirical basis for any choice — which otherwise could be completely arbitrary.

Finally, if blocking is viewed as a markedness condition, and if one accepts Carstairs' (1987) contention that free variations of inflexional suffixes, such as the variation between 3PL IND PRES /ε/ and /ny/ in Spontin Walloon, also are marked (and tend to be either reinterpreted or lost), then there are no reasons not to include this restriction in the formulation of blocking. In particular, principle (i) should be revised to avoid the application of two equally specific rules.