FII is now on Telegram
10 mins read

Posted by A Mani

Sexism is prejudice or discrimination based on sex or gender, especially against women and girls. The term is often understood, in these times, to mean prejudice or discrimination by patriarchy primarily against women and people of other genders. Specifically the term ‘cis-sexism’ is used to mean sexism based on fixed patriarchal ideas of gender binaries.

The first National Science Day was celebrated in 1987. Past themes have revolved around topics of immediate concern in STEM that require remediation. This year the theme is ‘Women and Science’. An acknowledged fact within this domain is that women are still few in the higher echelons of science and most women have testified to the sexism that exists in the scientific community. There are multiple studies that reflect the gender-stereotypes that exist in the scientific workspace. There are also studies that show it’s not just the ‘scientific’ workplace that sustains sexism, the work too deserves some attention on how sexist views seem often neglected and are even passed on as par for the course.

Image Source: PLOS Blogs

Issues relating to these have been recognized by a number of scientists, and policy makers. But a substantial number of these people still live in denial. According to the available literature on science, till recently (before a few years ago), the structure of the female clitoris, the fact that women feel pain differently from men, the fact that an ovary takes about fifteen minutes (as opposed to less than 15 Secs) to release an egg cell, and other research pieces on trans women were not known earlier because of systemic sexism. Sexism has been entrenched in academic activities for centuries due to oppressive religious, patriarchal power structures in most countries in the world. Therefore, people who are brainwashed by such power structures tend to oppose policy makers and scientists that seek to create a relatively equal, non-discriminatory world.

There are multiple studies that reflect the gender-stereotypes that exist in the scientific workspace. There are also studies that show it’s not just the ‘scientific’ workplace that sustains sexism, the work too deserves some attention on how sexist views seem often neglected and are even passed on as par for the course.

Scientific work is theoretical and experimental. In 2020, I should not be saying that women bring their own precious sensibilities to the table and that the biases created so far should be seen as gaps in or absence of knowledge that are waiting to be corrected. Any corrective action that can be taken in the context of the problem depends on a complex maze of patriarchal, social, religious, political and economic factors.

The practice of science has a number of avenues for improvement and many that require corrective action in retrospect. Among these, sexism in the formalism and performance of scientific experiments is often subtle or not so obvious. In the last few decades, more attention has been devoted to these by at least some researchers and much has been discovered about its magnitude and colossal impact on women in general. But they are yet to make substantial impact on policy and practice at a global level as of this writing.  

Image Source: Yahoo

While anything that affects the active participation of women in STEM is bound to affect their active participation in scientific experiments, the converse is relatively more complex. We can expect motherhood myths such as those relating to the retrograde hetero-sexist view that women’s work threatens children and family life to adversely affect sciences that involve active experimentation. This can be seen in the indicators mentioned in a large sample study spread over 18 countries. Similar patriarchal views are pretty common across religions and caste in India and sustained struggles are necessary for improving the situation.

Some scientific experiments are real, some less real and others are thought experiments relative to the level of perception involved in conducting the experiment. Some examples of a scientific experiment are 

  1. a sociological study that involves statistical sampling; 
  2. a medical drug testing trial with control;
  3. a mechanics experiment to confirm theoretical predictions in rotational kinematics;
  4. drug testing on animals;
  5. a sociological study on the impact of divorce on divorced women;
  6. an unsupervised machine learning algorithm;
  7. a big data algorithm working in the context of data from human sources;
  8. a supervised machine learning algorithm working on big data; and 
  9. an archaeological study of graves of ancient people.

Interestingly, some experiments may not be properly motivated or may be motivated by non-scientific reasons that may be sexist in addition.

Not really beginning with Darwin’s ideas about women as ‘under-evolved’, scientific experiments have managed to either ignore women or left them out for some ‘unscientific’ reason. Darwin did contribute a lot to the study of evolution of organisms, but his unscientific utterances had considerable effect on the rest of the male population.

Not really beginning with Darwin’s ideas about women as ‘under-evolved’, scientific experiments have managed to either ignore women or left them out for some ‘unscientific’ reason. Darwin did contribute a lot to the study of evolution of organisms, but his unscientific utterances had considerable effect on the rest of the male population. It took Eliza Burt Gamble to dispute his take on the status of evolution of women; she educated herself and set about to prove him wrong. Helen Hamilton Gardner proved wrong the idea that women had smaller brains and were stupid. 

Also read: How Psychology Wronged Women

Researchers have pointed to the imbalance created by deliberate exclusion of female subjects in experiments in neurosciences. The authors reviewed reports of experiments that call for inclusion of sex as a biological variable, with specific reference to experiments studying MIA (maternal immune activation) model in humans. Experiments had either analyzed only male data or either used male and female data and not reported the differences or analyzed male and female but reported that no differences exist without providing analysis.

Image Source: The Tartan

One of the reasons cited most often is the historic development of paradigms but these deliberate selections may also stem from certain hard-wired neglect. One of the specific examples that researchers discuss while illustrating the importance of properly designed experiments is that of MIA challenged rats producing female and male offspring. The female offspring had higher levels of IL-6 mRNA (responsible for inflammation and maturation of B cells) while male offspring rats showed only spatial impairment. Such findings need further exploration among human cohorts as these may have strong implications in health and disease studies. The sex selection or neglect has becomes a serious concern.

Studies in psychology that associate certain topics with gender stereotypes reinforce those stereotypical definitions and also influence their outcomes. Studies may also show bias from constraining analyses with known theories and hence limiting our understanding. Examples of studies on behaviors of women and girls might fall into such traps of in-deliberate ignorance. Some recognition of the problems have been in the air for over half a century as evidenced by an early article on biases in psychology. Psychological aspects of the general issue in STEM are reviewed in the collection.

Image Source: Undark Magazine

Many science experiments involve field work of different types. In an alarmingly large number of these women researchers have experienced sexism and sexual harassment. In particular, overt sexist behavior is not uncommon in archaeology field work. The following quote from the archaeology sucks blog (arch) says a lot,

…but I’ve seen male coworkers treat female coworkers with disrespect, causing me to instantly lose all respect for those guys, and feel uncomfortable whenever I had to work directly with them. Male crew chiefs who made a game out of trying to make female field techs cry at work. Male coworkers openly discussing the comparative sexual appeal of their female coworkers. Male coworkers casually exposing themselves, both at work and in social settings.

Many scientific experiments involve field work of different types. In an alarmingly large number of these women researchers have experienced sexism and sexual harassment. In particular, overt sexist behavior is not uncommon in archaeology field work.

Post Doc Alex Jones had this to say in 2014 (when she was working in the Lyminge Project in UK),

It is surprising how many times I have discussed with a female student what I expect them to do and have had them protest that they aren’t able or will be slower because they are ‘a girl’. Contrarily, it also surprises me how annoyed some of our male students are at seeing a female archaeologist out-perform them (I must emphasize that this indignation is rare, but has happened).

Sexism in Artificial Intelligence and Machine Learning

Way back in 1973, part of Playboy center-fold model, Lena was first used in image recognition studies. The sexist practice has continued since in the field of machine learning with its not so subliminal sexist messages. Only in recent times has the practice been stopped in some universities.

In supervised learning, algorithms are trained by the data scientist on a subset of the data, then tested on another subset and further validated on yet another part of the data. The training data set may include many instances of sexism that may result in the algorithm becoming sexist. If nine out of ten women in the training set had been viewed as incompetent by male bosses for sexist reasons, then the algorithm is more likely to assume that women are incompetent. Such sexism may be rectified by a scientist with proper training in feminist perspectives during the training process – the person in question should be able flag or adjust the labels in data set appropriately.

Image Source: Foreign Policy

When the data set becomes too big with respect to number of attributes and objects, then a reasonable training set is also likely to be very big in size. Labeling of such training sets may be very sketchy. In this big data scenario, algorithms are likely to learn most sexism inherent in the data. Effective solutions to this problem are not known as of this writing. Any method of addressing this can only be through relatively ontology-sensitive sub-processes.  

All this suggests that the best way forward would be integrate feminist theories in a functional way into contextual decision making process at both personal and machine levels. By ‘contextual’, I mean at a suitable level of abstraction. In relation to scientific experiments, the source of sexism can be specified through a classification of agents, and effective combat deployed by the same.

Unsupervised algorithms do not involve intervention from the user and such algorithms are even more likely to learn sexist nonsense (whenever they are present in the data). Contextual or semi-supervised (unsupervised with a layer of supervision) are likely to be more optimal for these contexts. This scenario is because computers are not learning anything, are not thinking, and are stupid. A number of papers have been published on these issues and a lot remains to be done.

All this suggests that the best way forward would be integrate feminist theories in a functional way into contextual decision making process in science at both personal and machine levels. By ‘contextual’, I mean at a suitable level of abstraction. In relation to scientific experiments, the source of sexism can be specified through a classification of agents, and effective combat deployed by the same.

Aspects of a Possible Classification Strategy

While one can engage endlessly in the quest to convert wild jack asses into oak trees through protracted engagement, there is reason for disbelief. Modern applied feminism speaks differently. If these are the problems, then there are some possible solutions based on certain minimal assumptions. Some scientists live in denial, believe in the religious patriarchy, have no clear commitments to the methods of science, and may have got into the profession by way of deficiencies in its recruitment process (such as nepotism and improper imposition of standards). Some others may not bother about conflicts between their own sexist religious and scientific beliefs because they exist as real contradictions. 

Also read: How Inclusive Is STEM In India? 15 IITs Have No ST Faculty

Still others may hold rational, egalitarian beliefs, understand the methods of science and the need for feminist intervention in their work environment. But may be constrained by peer pressure or a commitment to people pleasing. A number of variants of these beliefs are possible and it is not too hard (can be a bit cumbersome) to define classifications based on the cues. The main advantage would be that it would be easier to use targeted feminist discourses on them. Relative to the methods used by feminism-informed sites, the suggested strategies would have greater scope for engagement eventually. 

References

  1. Archaeology Sucks Blog
  2. Conceptualizing Brahmanical Patriarchy in Early India: Gender, Caste, Class and State by U. Chakravarti
  3. Sex and gender bias in the experimental neurosciences: the case of the maternal immune activation model by P. Coiro and D. D. Pollak
  4. It’s Time to Retire Lena by C. Culnane and K. Leins
  5. Guidelines for avoiding sexism in psychological research: A report of the Ad Hoc Committee on Nonsexist Research by F. Denmark, N. F. Russo, I. H. Frieze, J. A. Sechzer
  6. The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World by P. Domingos
  7. What science has gotten wrong by ignoring women by C. Zuckerman
  8. Justifying gender discrimination in the workplace: The mediating role of motherhood myths by C. Verniers and J. Vala
  9. My experiences in the field by A. Jonas
  10. Than One Swallow to Make a Summer: Measures to Foster Girls’ and Women’s Pathways Into STEM
    Overview of Quality of Life of Older Lesbians and Trans Women in India by A. Mani
  11. Nature
  12. Gender Stereotypes in Educational Software for Young Children by J. P. Sheldon
  13. Women’s representation in science predicts national gender-science stereotypes: Evidence from 66 nations by D. I. Miller, A. H. Eagly, M. C. Linn
  14. Gendered Paths into STEM. Disparities Between Females and Males in STEM Over the Life-Span by Bernhard Ertl Silke Luttenberger Rebecca Lazarides M. Gail Jones and Manuela Paechter
  15. Older and Younger Adults’ Attitudes Toward Feminism: The Influence of Religiosity, Political Orientation, Gender, Education, and Family Sex Roles by K. E. Fitzpatrick Bettencourt, T. Vacha-Haase, Z. Byrne

The author would like to thank Reema (Dr Reema Mani of HBCSE) for useful remarks during the process of writing this article.

A. Mani is a leading researcher on the foundations of rough sets, algebra, logic, rough sets, vagueness, mereology and foundations of Mathematics.  She has published extensively in her areas of research in international journals. She is a senior member (elected) of the International Rough Set Society, a visiting faculty/researcher of HBCSE, TIFR Mumbai. Mani is also an active feminist, lesbian rights and free software activist. She has been involved in few projects and has published a number of academic and popular articles on these subjects as well. You can find her on her Website or Orcid.

Featured Image Source: NPR

Support us

2 COMMENTS

Leave a Reply