122222 SE Linguistics Seminar / BA Paper (2023S)

In contrast to biological sex, gender is a social construct, and it is constructed verbally. Language use, as reflected in corpora, reveals significant asymmetries in the ways in which genders are constructed. For example, the verb _to kill_ is much more likely to take _he_ as a subject than _she_, while _she_ is much more likely to be the subject of the verb _to dance_. Likewise, the possessive pronoun _her_ is much more likely than _he_ to be followed by the noun _children_, although every child has both a mother and a father. On the other hand, the noun _followers_ is much more likely to be preceded by _his_ than by _her_. In fact, there seem to be very few words use usage does not reveal biased conceptualisations of gender roles.

At the same time, corpus evidence also reveals that usage has become significantly less gender-biased during the last 200 years. For example, the nouns _colleague_ and _job_ were hardly ever preceded by _her_ in the nineteenth century. Since the 1960ies, however, _her job_, and _her colleagues_ are about as common as _his job_, and _his colleagues_.

In this course, we investigate the development of gender bias in large diachronic corpora.

Our primary goal is to quantify changes in gender bias during the last two hundred years. Therefore, the first part of the course will offer an introduction to the use of diachronic corpora, and to the basics of statistical analysis in Spreadsheets like Excel.

At the same time, however, we shall also look at our data also qualitatively, in order to avoid being misled by the evidence of mere numbers. For example, the mere frequencies of the phrases _he manages_ and _she manages_ may tell us very little, unless we know what it is that is managed by him or by her.

Also, we shall have to discuss what asymmetric frequency distributions mean in each particular case. For example, the fact that the phrase _working mom_ is more frequent than the phrase _working dad_ does certainly not imply that it is considered more normal for mothers than for fathers to be working. Rather, the opposite is more likely.

Thus, the goal of this course is twofold: on the one hand, we try to find out if language use has become more or less gender-biased over time, and on the other hand, we try to find out if corpus evidence is useful for addressing this question.