Two syntactic positions of grammatical gender

Abstract

The paper proposes some formal and functional criteria for distinguishing between two different syntactic positions of grammatical gender: determiner gender (D-gender) and nominal gender (n-gender). Focusing on D-gender and how it differs from n-gender, this work supports previous analyses of gender as a heterogeneous category that occupies different positions in the syntactic tree. Data are presented from 27 languages, many of which are either critically endangered or already extinct.1

Gender ‘Translation’ And Distributed Gender: Evidence From The Norwegian DP And Language Mixing*

“…As we will briefly come back to towards the end of this subsection, Norwegian seems to exhibit a "fixed" ngender system in Steriopolo's sense. Kucerov a (2018), similarly to Steriopolo (2018), proposes what may be called a distributed gender type analysis based on Italian data where gender features within DP are either valued from the lexicon or from the context. She adopts a simple model with just D+n+root with gender features on D and n. Then, if gender is valued in the lexicon, n starts out with a given gender value, and the unvalued gender feature on D is valued by an agreement operation with the lexically valued gender feature on n (Agree).…”

“…As we have shown, although gender may be flexible in Norwegian DPs, it is a distributional flexibility that is not correlated with semantics and/or pragmatics, and thus the various occurrences of gender in the Norwegian DP are semantically/ pragmatically uniform. Specifically, Norwegian DPs do not have discourse conditioned gender ("D-gender") in Steriopolo's (2018) sense, but only exhibit a "n-gender" system which is typical of Indo-European languages according to Steriopolo (2018).…”

“…Specifically, big or tall objects tend to be masculine, whereas small or short objects tend to be feminine. For instance, sebox‐r ‘table‐ m ’ refers to a high table, whereas sebox‐u ‘table‐ f ’ refers to a squat table (Steriopolo 2018:126). We take phenomena like those from Amharic and Sare as support for our assumption that gender‐related semantic/conceptual properties are translated into formal gender values by the gender translation mechanism.…”

“…Thus, on their analysis gender is not flexible in our sense, but it is still a distributed gender analysis in the sense that the different types of gender that they assume, are located on designated heads on the DP spine. Relatedly, Steriopolo (2018) distinguishes between “fixed” n‐gender and “variable” D‐gender, where n‐gender is lexically fixed and can take part in “fixed” gender agreement relations up the DP spine. On the other hand, D‐gender is discourse dependent and can be assigned depending on the contextual situation.…”

Gender as a property of words and as a property of structures

“…Within and alongside the discussions of 'semantic' and 'morphological' gender briefly referenced in 2.3, a strand of morphosyntactic research focuses on DPs where D has a semantically motivated gender value different from that of the head noun (beside the Russian data analysed by Steriopolo & Wiltschko 2010and Pesetsky 2013, Panagiotidis 2019. This falls under the rubric of gender determining different readings according to where it is marked in DP, not just in terms of 'high' and 'low' (Steriopolo & Wiltschko 2010;Steriopolo 2018) but also on intermediate loci along the DP spine. A prominent example is Fassi-Fehri's (2018) analysis of feminine marking in Arabic.…”