Motivating the (non-random) distribution of plural feminine pronouns in Hebrew The role of Humanness

While normative Hebrew distinguishes between atem/hem ‘you/they-msc-plural’ and aten/hen ‘you/they-fem-plural’, many current speakers use the masculine forms for both feminine and masculine plural referents, and it has been claimed that the masculine/feminine distinction has disappeared from the spoken language (Borochovsky Bar-Aba 2002). However, corpus examination shows many cases where the same speaker alternates between the masculine and the feminine forms when referring to feminine plural nouns. I therefore argue that the feminine pronouns are indeed in the process of being eliminated, yet they are not extinct. The purpose of this work is twofold – first, I will explain why the feminine plural pronouns are being lost, and second, I will show that the current distribution is not arbitrary, but rather affected by motivations other than gender marking, which are in line with universal typological tendencies. I adopt a usage-based approach to language change, and show why the fact that the feminine forms are giving way to the masculine forms is well-motivated. The main driving factor in the process is the low frequency of references to females in general and to plural females in particular (Ariel 2000). In addition, since mixed gender groups are referred to by the masculine pronouns, speakers have even fewer opportunities to use plural feminine pronouns. Furthermore, the feminine and masculine forms are distinguished only by their final nasal consonant (atem versus aten hem versus hen). Thus, the phonetic similarity may be seen as a facilitating factor, and it can explain a puzzling fact – why the loss of gender distinction in Hebrew is restricted to the pronominal system, when cross-linguistically, pronouns are the category most likely to retain distinctions lost elsewhere (Corbett 1991). Next, I claim that the current distribution is not arbitrary, and I demonstrate this by showing its sensitivity to the Animacy hierarchy. I have conducted a picture-description experiment, in which the subjects used pronouns that refer to several different plural entities that are grammatically feminine. The results indicate that speakers are significantly more likely to use feminine pronouns when referring to human, than to either inanimate or non-human animate referents. I argue that this distribution is also well-motivated. Even in languages with formal gender systems, such as Hebrew, pronouns are shown to be affected by semantic gender ( Corbett 1991), and psycholinguistic evidence demonstrates that animate and inanimate nouns are not treated in the same way (Deutsch et al, 1999). If so, using the masculine pronoun for female humans not only violates normative grammatical gender, it also violates natural semantic gender. This motivates why inanimate nouns are more susceptible to the loss of gender distinction. The fact that only humans, and not all animates, exhibit this preference, can be explained by the fact that pronouns which refer to humans are overwhelmingly more common than pronouns which refer to animals.