Final report

Summary of final report

Aims and objectives
The Project aimed to provide a comprehensive comparative-historical description and explanation of the morphological marking of singular and plural in the Romance languages, with a publicly accessible online Database of forms.

Findings
Beyond enriching our knowledge of typology of plural-marking across Romance languages, our research has thrown new, often unexpected, light, on issues particularly involving the fate of remnants of Latin neuter-plural markers in Romance and referential non-identity between singular and plural.

1. The nature and origins of the genus alternans (whereby nouns have a different gender in plural and singular). This proves to be dependent purely on morphological form than has hitherto been admitted, which leads us to challenge most established accounts of the phenomenon.

2. Romansh dialects have masculine nouns whose plurals denote 'collectivities' and which express plurality via a feminine singular form. Our work underscores that these 'collective' feminine singulars display plurals of their own, emerging particularly in expressions where the meaning is figurative and in the context of lists whose other items are morphological plurals. Future research will involve fieldwork and further exploration of corpora to ascertain the precise contexts in which ‘plurals of plurals’ arise.

3. Semantic ‘distance’ (referential non-identity) between singular and plural has been a constant interest. A spectacular effect of such non-identity arises in Romanian, where analogical extension of the plural desinence -uri from the genus alternans into feminine nouns, thereby destroying an established and inviolable pattern of morphological organization in feminines, proves exquisitely sensitive to referential non-identity between singular and plural. This observation supports Maiden’s claim that crucial in survival of systematic distributionally idiosyncratic patterns is perfect synonymy between members of inflexional paradigms.

4. ‘Suppletion’(that phenomenon whereby what appear to be different words figure in different parts of some paradigm — cf. go - went) is standardly assumed not to exist in Romance nouns, let alone adjectives. We find suppletion precisely in certain adjectives, and we are able to show how the interaction of the meaning of these adjectives with the meaning ‘plural’ generates abnormal behaviour.

5. Ampezzan (Ladin) reveals ‘overabundance’, multiple plural forms corresponding to one form for the singular. This argues against the notion that plural forms tend to be stored lexically, for they seem more readily compatible with ‘rule competition’, such that more than one pluralization rule can be applied to selected lexemes having a particular phonological shape.

Type of publication resulting
We plan to publish at least 8 articles, a book and an online Database. We have made, or will make, at least 21 conference presentations.

Strengths and weaknesses
We have attained original insights into the evolution of remnants of Latin neuter plural marking in Romance, requiring major revision of established views. We have been less successful in identifying the role of lexical semantics in determining irregularity in plural formation. We were probably over-ambitious in believing we could analyse the lexical semantic determinants of irregularity for individual nouns in the time available. This aspect of our research will yield results in the longer term.