Data Ethics Club: Statistics, Eugenics and Me#
This is summary of Wednesday 17th November’s Data Ethics Club discussion, where we spoke about the article Statistics, Eugenics and Me by Raphael Sonabend.
The summary was written by Huw Day, who tried to synthesise everyone’s contributions to this document and the discussion. “We” = “someone at Data Ethics Club”. Nina Di Cara and Natalie Thurlby helped with a final edit.
The article outlines the experiences of the author as they studied in the Statistics department of University College London (UCL), which was the home, for decades, of the most prominent eugenicists. This was juxtaposed with the author having had many relatives murdered in the holocaust. As the author puts it:
“Much of my family was murdered in the holocaust as part of a regime to eradicate a supposed ‘inferior race’. Yet, I still viewed Pearson, Fisher, and Galton (and others) as the Fathers of Statistics who deserved to be recognised and respected for their contributions. I had naïvely assumed they were a product of their time and that their research was a natural progression in the statistics behind genetics. This was not the case.”
In this discussion, we considered the history of the relationship between statistics and eugenics, our awareness of this relationship (including how much it is talked about in formal education settings) and the potential pitfalls with hero worship both within and outside of academia; plaques, statues and buildings named after particular scientists.
To what extent were you already aware of the relationship between statistics and eugenics?#
We had a mixed discussion involving data scientists, genetic epidemiologists and philosophers as well as those with more general STEM backgrounds (physicists, mathematicians, engineers etc.) and the response to this question was highly dependent on field. Philosophers and historians consider the link well-known, but may have been suprised to learn that many mathematicians and scientists were hardly (if at all) made aware durung their statistics lectures - and certainly not earlier (in GCSEs or A-levels). We discussed how this is due to a teaching focus on tools and how to use them within many STEM modules (sometimes served alongside a surface-level quirky fact), as well as the tradition of results being attributed to their creator (e.g. Euler’s Theorem). There was also variability within the sciences, for example those with a background in genetic epidemiology knew a bit more about the links with eugenics.
One group member realised that they had recently seen a blue plaque commemorating Fischer on the gate-pier of Inverforth House, North End Way, Hampstead, where he lived as a child between 1896 and 1904. “Sir RONALD AYLMER FISHER 1890-1962 Statistician and Geneticist lived here 1896-1904”. There is no mention of eugenics in the (admittedly small) plaque. Should his pivotal role in the field of eugenics be acknowledged?
We mused which parts of statistics would be changed by these views and whether interpretations of statistical results might have been reframed by the beliefs of some statisticians? Would statistical tools have been developed in the same way if they were being applied to different problems? Or would the scientific culture around when to apply these tools to be the difference (if any)?
How do you think that a researcher’s support of racist or ablist eugenics should impact the way their work is taught, or how their names are used for other purposes in research communities?#
We asked oursleves: “Who deserves to be remembered?” in the context of statistics (and scientific discoveries more generally). Could other people have discovered p-values (for example) and not been a eugenist? Was it the motivation of studying eugenics that lead to these tools being developed? Many of us who had studied STEM subjects in higher education noticed a culture of objectivity, where it is seen as wrong and not important to delve into the who, what, why? There is a separation of statistics from its social context in STEM education.
Do we stop using the tools just because someone racist made it? No. We need to separate the work from the people whilst still acknowledging their problematic views. Finding ways to use statistical tools/methods in ways that counter eugentics/racism , that is, using them as tools for justice, would be a way to honour the accomplishments of the methods, and not the individuals who devised them.
One (extreme?) suggestion would be renaming the methods. One member of the group reminded us that Jimmy Saville had several statues in his honor, and as this letter notes: “No one would argue that removing statues of the sex criminal is ‘cancelling culture’, yet this logic is routinely used to defend monuments of slavers.” But are we honouring a person by naming a method after them, or simply acknowledging who first documented it? Indeed, this approach may be attempting to forget/disregard the past rather than confront it. Instead the general consensus was that we should continue this idea of acknowledging someone’s contribution to the field, regardless of their problematic ideas, whilst acknowledging and talking about those ideas. What ideas did they have? Why are they problemtic? How did those ideas potentially contribute to their innovations in this field? How can these innovations be used for good or bad?
An example from philosophy is Martin Heideger, who is best known for contributions to phenomenology, hermeneutics, and existentialism. He was also a member of the Nazi Party and there is controversy as to the relationship between his philosophy and his Nazism. This is an ongoing discussion of how much his views affected his philosophy.
What do we make of the defense “they were a product of their times”? That you should still be able to call someone out, even if their ideas were a product of their culture. This does not rule out them having other, good ideas. Should a Professor who has discriminatory views be allowed to hold their post? In the past, yes. Nowadays, less so. We discussed previously about wanting to decolonise academia and the push for more diversity in academia. Linking in to our previous discussion on decolonising computer science, how can we hope for more diverse individuals to progress through academia if the very people they look up don’t want them there?
How accountable should people be for their opinions? Free speech is important particularly to an intellectual community, and most of us felt uncomfortable with it being policed, but that doesn’t mean people shouldn’t expect consequences to the things they say and how they act. If you called someone a racist slur on twitter, you would expect to be labelled a racist. Some felt that being held accountable isn’t a credible threat to free speech; as the saying goes “If we can see you complaining about being cancelled, you haven’t been cancelled”. However, anyone with an opinion (or simply a public identity) online can be subject to abuse or simply pile-ons of people disagreeing, which can lead to them removing themselves from platforms. We didn’t come up with a quick and simple recipe for detecting unacceptable speech intended to harm, and we are aware that lazy attempts to moderate can go wrong.
Should your opinions cost you your job? We know that they can, but do we agree? If your opinion is denying the existence of some of your students (in the case of transphobia), then should you be allowed to teach them and express that opinion? Most of us thought “no”, because it would create an unwelcome atmosphere for learners, much like the author of this piece was learning statistcs in a building named after someone who didn’t want them to exist. This may then continue to bias who stays in research or other fields of work. In other situations, however (e.g. most Data Ethics Club discussions), diversity of opinions can be useful: echo chambers can be harmful, and being able to have a lively debate about something can be useful and simply more interesting as long as the topic of discussion doesn’t ostracise members of the workplace/community.
What change would you like to see on the basis of this piece? Who has the power to make that change?#
Based on this piece, we would like to see teaching on social context included/embedded it learning scientific methods. This will be hard in practice, both to train teaching staff and to embed interdisciplinary work in STEM subjects, but those in teaching positions do have the power to do something about this, but creating decolonised materials doesn’t need to be solely the work of lecturers. They would need to be supported to do this. In the School of Politics and International Relations at the University of Bristol, there is student-led activism to decolonise the curriculum, resulting in ranking the teaching with a traffic light system by ‘white lens’. Following that, the school of Biological Sciences is now seeking to hire a PostDoc to begin the work of decolonising that curriculum.
We also need to educate ourselves on the histories of our own fields since we didn’t get that education. One attenddee said: to work in the field of data science is to work in the shadow of eugenics. Acknowledging and learning from this could better inform how we go about innovations in the future, as well as reminding us of the imperfections of various academic role models.
Another member made the point that naming buildings after scientists/making statues of them fuels the “lone rockstar” view of scientists. Whilst in the past, this idea might have been more prevelant, by today’s standards, it encourages a view that is unrealistic (and arguably incompatible) for the next generation of scientists and academics. Therefore, creating a culture of Team Science is also on our wish list.
Huw Day, PhDoer, University of Bristol, @disco_huw
Euan Bennet, Senior Research Associate, University of Bristol, @DrEuanBennet
Sergio Araujo-Estrada, Aerospace Engineering, University of Bristol
Laura Sheppard, PhD Student at CASA, UCL @Laurahsheppard
Giulio Centorame, Visiting PhD student & Research Associate, University of Bristol, @GiulioCentorame
Vanessa Hanschke, PhD Interactive Artificial Intelligence, University of Bristol
Roman Shkunov, Maths/CS student, University of Bristol,@RShkunov
Kamilla ‘Milli’ Wells, Citizen Developer
Mia Mace mace-space