Volume 45, Issue 2 - The Diachrony of Gender Marking

This article addresses various issues in the diachrony of gender marking, such as the origin and typology of gender systems, pathways of change and the question of directionality in relation to the Agreement Hierarchy, and the semantic basis of changes in gender systems in relation to the Individuation Hierarchy. It also offers an overview of recent multidisciplinary approaches to the evolution of gender systems including language acquisition research, contact linguistics, and theoretical syntax.

The Dutch pronominal gender system is believed to be undergoing a process of resemanticization. While most research has focused on northern varieties of Dutch which are ahead in the change, this article describes the ongoing demise of the traditional three-gender system and the incipient resemanticization of pronominal gender in a southern variety of Dutch. This process can be considered an instance of morphological regularization – that is, the rise of an innovative rule system when the traditional system becomes too opaque to be successfully acquired. Resemanticization takes effect along cross-linguistically widely attested lines (the Agreement Hierarchy), using parameters that are typologically common in gender systems, such as animacy or individuation. It is argued that the locus of language change is the language acquisition process: indeed deviations from grammatical gender in adults are typically in line with the semantic system for pronominal reference that is found in young children. As children grow older and their lexicon expands, these semantic rules which categorically determine their pronominal reference at the age of three are turned into default rules, which mainly apply in cases where there is uncertainty concerning a noun's grammatical gender, for example, for infrequent nouns.

The gender system of Irish appears to have undergone a process of simplification: traditionally depending on both formal and semantic assignment rules, agreement in contemporary spoken Irish is still rather conservative within the noun phrase, but almost exclusively semantic anaphorically. Language contact and the resulting obsolescence seem to have had some influence on these developments: for instance, structures that have a functional counterpart in English seem more resilient than others. But language-internal developments, particularly the phonetic erosion and loss of word-final syllables, may have played an important role, too: similar developments have been observed in non-obsolescent languages like Dutch and French. In this article, I illustrate some specific aspects of the Irish situation with examples drawn from a corpus of spoken Irish and frame the simplification process in terms of structural convergence in the context of language contact.

There are many examples of gender loss in the Germanic languages, but none of the rise or return of a particular gender. However, there has been substantial change in gender morphology, involving an increase in morphological complexity. The gender system of German appears to be extremely robust when it comes to the number of genders. In this article, I will introduce a principle for the semantic restructuring of gender systems. Focusing on target gender morphology, I argue that constructive changes have as their point of departure semantic features that lie at the core of the gender system. I stress that such changes affect sub-systems, resulting in layers of gender markers that can evolve at different paces. To exemplify how layers of sub-systems can serve to uphold an overall tripartite gender system, I use the repeated restructuring of semantic distinctions in Germanic adjective inflection.

In the Germanic languages, gender and declension are two classification systems with a restricted functional load. Still, both persist in many languages, and in some of these languages they are even intimately interrelated, i.e. gender can be predicted based on declension, or declension can be predicted based on gender. Several Germanic languages and dialectal varieties of German are compared with respect to this link between gender and declension. Based on contrastive data, the interaction seems to depend on the level of complexity, i.e. the number of declensions and genders. When complexity decreases, the conditioning of both categorization systems is either more strongly interrelated (leading to parallelization in the most extreme cases), or gender and declension are dissociated and bound to new, more transparent conditioning factors. The developments are interpreted against the background of the hypothesis that gender and declension are used complementarily in profiling the number category: gender profiles the singular, whereas declension profiles the plural.

A long-established tenet of Indo-European linguistics says that grammatical gender systems all along the history of this language phylum were maximally tripartite and generally tended toward a reduction of gender contrasts. In this article, we shall show that this widely-held idea overlooks the existence of four-gender systems in a substantial part of the Romance language family, a fact that has in turn gone unnoticed so far. We shall provide an analysis of the relevant Romance data, a sketchy comparison with other four-gender systems described in linguistic typological research, and a detailed reconstruction of how the gender systems in question might have developed in the Latin-Romance transition.

Gender fulfills two different functions, i.e. nominal classification and crossreference of constituents through agreement. Besides the generally acknowledged possibility of a grammaticalization process that may lead classifiers to become gender markers, gender systems may also arise as a consequence of special agreement patterns connected with differential marking of core arguments. It is argued that different origins of gender systems imply higher relevance of either function of gender in individual languages, and that this may have consequences on the values of gender within specific gender systems.

This article analyzes gender-agreement variation in the Determiner Phrase (DP) of Afro-Bolivian Spanish (ABS). We claim that this is an instance of crossgenerational change, consisting in the systematic substitution of stigmatized basilectal Afro-Bolivian features with more prestigious Bolivian Spanish ones. In light of recent minimalist models, the variability encountered in the corpus can be accounted for systematically as a by-product of the differential specification of uninterpretable features in a derivation. This study highlights an on-going process of post-bozal Spanish approximation to a more prestigious Spanish variety. Such transition seems to be driven by social factors and is significantly regulated by syntactic constraints.

In this study we discuss the distribution of gender exponents in Old and early Middle English based on an analysis of two historical documents (Orosius, Peterborough Chronicle). The gender exponents investigated include demonstratives, adjectives, numerals and pronouns. We analyzed 179 noun phrases from the Orosius and 1,090 noun phrases from the Peterborough Chronicle. While the Orosius illustrates a highly consistent distribution of gender exponents, the Peterborough Chronicle contains substantial variation. As for the Peterborough Chronicle, we can demonstrate that the number of gender exponents that is used in conflict to the Old English gender system increases over time. In addition, we investigate the most important factors responsible for the observable variation focusing on the properties of the head nouns and their referents. Our results show that noun phrase internal and noun phrase external gender exponents behave differently. Moreover, formal properties of the head noun (structural and morphological case, number) are better predictors for gender variation than the properties of the referent (abstractness, degree of individuation).