Cultural connotations of categorizing the environment: does the presence of a linguistic gender and noun class system in any way connect to cultural feature data?

Social studies indicate that a gendered language may limit equal opportunities for women (Jakiela and Ozier 2018). Likewise, the use of gender-neutral pronouns may improve gender equality (Tavits and Pérez 2019). We aim to investigate this issue using cross-cultural data. The studies of (Whyte 1978) and (Sanderson and Donoghue 1989) (W1978/SD1989) connect gender inequality to cultural features.
Testing three theories, the Warfare hypothesis, the Marxian hypothesis, and the Non-marxian materialist hypothesis, they found significant effects only for the latter (lower percentage of contribution to food by women, intense agriculture, use of plow, patrilineality, partilocality). Wh1978/SD1989 used a sample of 186 cultures, selected to avoid Galton effects. We use a global set of linguistic gender/noun class (3079 languages), retrieved by automation (Virk et al. 2017) and corrected manually. We extracted the features of Wh1989/SD1989 from D-PLACE (Kirby et al. 2016). We tested (using a mixed model) the inequality features Domestic authority of women, Ritualized female solidarity, and Control of women’s sexuality, against linguistic gender and/or noun class. We found no correlation. We tested the features significantly correlated with gender inequality in W1989/SD1989 and found effects for noun class, which is a Galton effect (most noun classes are found in Africa). When we merged gender/noun class and tested against the significant features, we found several negative and positive correlations, connected to, e.g., participation in agriculture, crosscousin marriage, patrilocal residence, and intense agriculture. Therefore, we suspect that gender/noun class may correlate with subsistence and kinship, to which inequality may be another side-effect.