This article investigates the evolution of bare nouns, used without a determiner, through the history of the French language. The loss of bare nouns is charted through calibrated corpora of non-fictional prose texts from the same genres and region, ranging from the 12th to the 19th century. The change is first completed with nouns in subject function, significantly advances with direct objects, and progresses with obliques. The extensive quantitative documentation demonstrates that the change is impacted by the syntactic function of the noun, along the Accessibility Hierarchy. The speculation is examined that the more accessible functions encourage expression of (definite) determiners, thus explaining the pattern of change.

While major restructurings and simplifications have been reported for gender systems of other Germanic languages in multiethnolectal speech, this article demonstrates that the three-way gender distinction of German is relatively stable among young speakers from an immigrant background. We investigate gender in a German multiethnolect based on a corpus of approximately 17 hours of spontaneous speech produced by 28 young speakers in Stuttgart (mainly from Turkish and Balkan background). German is not their second language, but (one of) their first language(s), which they have fully acquired from childhood. We show that the gender system does not show signs of reduction in the direction of a two-gender system, nor of wholesale loss. We also argue that the position of gender in the grammar is weakened by independent innovations, such as the frequent use of bare nouns in grammatical contexts where German requires a determiner. Another phenomenon that weakens the position of gender is the simplification of adjective-noun agreement and the emergence of a generalized gender-neutral suffix for prenominal adjectives (that is, schwa). The disappearance of gender and case marking in the adjective means that the grammatical category of gender is lost in Adj + N phrases (without a determiner).

Based on a discussion of correlations between syntactic position, prosodic cues, aspect and generic vs. non-generic interpretations, this paper substantiates that Danish Bare Plural count nouns (BPs) have a wider distribution than Bare Singular count nouns (BSs). BPs, unlike BSs, can occur in subject position, function as both generic and existential arguments, and appear with all aspectual verb classes. However, BPs and BSs expressing a non-generic, modificational meaning concur in object position of activity verbs and stative verbs with a possession relation implicature. These V+BP and V+BS structures, it is suggested, form a progressive continuum of three different subtypes of pseudo-incorporation (PI), namely (i) PI of BPs (low integration as in spise æbler ‘eat apples’), (ii) PI of type 1 BSs (medium integration as in male hus ‘paint house’), and (iii) PI of type 2 BSs (maximum integration as in spille violin ‘play violin’).

This paper assumes that the basic denotation of nouns can be that of kind or property and that the determiner system of a language is a direct consequence of this cross-linguistic variation. An analysis of how definiteness and specificity are marked across three languages with different determiner systems, namely, English, French and Mauritian Creole (MC), provides evidence of the co-relation between noun denotation and determiner system. Languages with kind denoting nouns (English and MC) admit bare nominal arguments, which are barred in French, whose nouns denote properties. However, English and MC differ in that English has an overt definite article, which is a lacking in MC. This null element requires licensing by an overt specificity marker in some syntactic environments. The English and MC definite articles are analyzed as operators that quantify over sets of kind denoting nouns, and they serve a different function from the French definite article, which is specified for number and selects properties.