Gender Linguistics: Of Asterisks and Samples

My contribution "Why gender-fair language is not fair" criticized the research on the reception of gender language.

The authors of a study quoted as an example then complained about “inaccuracies” and “insinuations”.

However, they ignore decisive objections.

Like many similar studies, the one cited suffers from a problematic sample, an experimental design unsuitable for the research question and the everyday irrelevance of the measured differences.

In this specific case, it is undisputed that the participants in the study were not students of gender linguistics, but were recruited via an online platform.

However, no random sample was provided, but people who had previously proactively registered there, i.e. showed an interest in participating in scientific studies, and who had selected their own study from a list of study descriptions.

Although such a procedure is convenient for researchers and therefore not unusual, it harbors a considerable risk of bias, especially for our topic.

This is confirmed by the socio-demographic characteristics of the sample, which, according to the raw data, is essentially made up of young people with a university entrance qualification.

These cannot represent the entire population either numerically or ideologically and certainly not in their relation to gender language and are confronted with the affirmation of gender stars and associated ideologies in universities like almost nowhere else on a daily basis.

Even those who claim to reject gender language or are not familiar with it have previously been socialized through organizational requirements and constraints at their university, are in contact with colleagues or fellow students who enjoy practicing it every day, and regularly witness relevant ins Extremely increased sensitivities,

Relevance of age and context

The fact that such a background of experience can influence the response of test persons to the subject of gender language and that relevant dispositions in this regard cannot be componentized via individual questions is quickly evident from a social-psychological point of view.

As is the fact that it is difficult to make meaningful statements about "educational effects" from a sample that only includes around one percent of people without a school diploma or with a secondary school diploma (proportion of the population: around one third).

The same applies to age: people over forty years, after all the majority of the population, are hardly ever found in the sample, although it is well known that their understanding of and of gender stars declines rapidly.

The authors consider age effects to be irrelevant,

The high relevance of context information for our understanding of meaning is also known.

Linguistic meanings, especially those relating to gender, arise from the possibility of an extended classification of facts.

This is the only way we can always easily distinguish brightness-generating components of lamps from pome fruit and a former Federal Chancellor.

The fact that important context information on gender determination is systematically cut off in many studies on gender language has often been criticized and is only suitable for clarifying “medium-sized” effects in the misunderstanding that builds on it.

However, this did not prevent the authors from presenting their subjects with downright bizarre example sentences such as these: “The motorists came out of the building.

There will always be different opinions on how full a glass has to be to be full.

What counts as a “moderate” effect for people who have a perfectly legitimate academic career exploring significant differences may, without any contradiction, be completely irrelevant in everyday life.

In any case, the measured effect sizes, even quite apart from the problem of their methodological foundation, are hardly suitable for the assertion often made by interested parties that the use of the generic masculine "primarily" makes one think of men.

If unsuitable samples are not even discussed and well-known deficiencies such as the lack of contextual knowledge are silently reproduced;

if, in view of the statistically "medium-sized" effects of gendering, the huge ones in the completely unproblematic understanding of generic forms are not worth mentioning without any context and the citation of further studies unilaterally imply that positive effects would emanate from gender language - then the reader may get the impression of bias arise

It's nice when the authors take a stand here in their reply.

This, at any rate, is granted to you with the apologies: from the

indeed, they were not surprised by gender language.

This only holds the potential for surprise for people who mistakenly consider an allegedly gender-fair language to be fair and still regard this error as scientific knowledge.