Towards decolonising computational sciences (20th Oct 2021)#

What’s this?

This is summary of Wednesday 20th October’s Data Ethics Club discussion, where we spoke and wrote about the paper Towards decolonising computational sciences, written by Abeba Birhane and Olivia Guest. 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 the final edit.

What does it mean to decolonise something?#

We discussed Towards decolonising computational sciences, a paper by Abeba Birhane and Olivia Guest. As the authors set out in the abstract:

“This article sets out our perspective on how to begin the journey of decolonising computational fields, such as data and cognitive sciences. We see this struggle as requiring two basic steps: a) realisation that the present-day system has inherited, and still enacts, hostile, conservative, and oppressive behaviours and principles towards women of colour (WoC); and b) rejection of the idea that centering individual people is a solution to system-level problems.”

We can see what solutions the paper offers for solving problems and use that as a starting point for what the paper means by decolonising academia.

There are systems of oppression and power imbalances in academia and we want to get rid of those systems so that everyone is equally welcome and people are systematically treated better. This is the prevailing power of colonialism – in terms of knowledge, especially in education and academia. Decolonisation requires a willingness to give up and share power.

Decolonising academia could involve addressing power structures that exist due to colonialism and confronting them. Perhaps it relates to post-colonial theory and frameworks.

Decolonising could mean to become independent, and also to undo the effects of colonisation. Decolonisation is huge with farreaching ramifications. It is everything about our systems - school, politics. This framework needs to be dismantled to strive towards decolonisation.

How can we start decolonising academia?#

Increasing diversity, particularly at the top#

One suggestion was that, as a starting point, we can aim to reach a point where systemic discrimination is eliminated. Then the only discrimination left is from racist people. Then we can just weed those people out. This would involve eliminating leftover structures of decision making from colonial times. We need to to encourage inclusion, equality, diversity whilst withdrawing from colonial structures.

We currently see 2 in 5 students being women at the undergraduate level and then only 1 in 20 at the PhD level, we see this problem materialising. This kind of disparity can be explained by a relatively small (3%) undervaluation over a woman’s career. This bias can be accidental (non-malicious), but still result in a structural problem. As a concrete example of this 2.1% of professors in the UK are BAME Women - 25 of +21,000 are black women. People get fussy about quotas, but could they be part of a solution in some cases? Or would simply recording people’s biases be enough to prevent their effects?

This leads us to question: can we address colonialism in academia without decolonising all of society? Should we approach this problem top-down or bottom-up? Either choice feels depressingly bound to take a lot of time. As an example, Australian education slowly shifted from cherry picking colonial pride moments to including full history and perspectives (arrival vs invasion).

Reducing pseudoscientific racism#

People who are vocally racist or misogynist can often be caught using a pseudo scientific mode of reasoning. Examples include: The Google memo, eugenics, etc. Pseudo science is inherently linked to discrimination, but this link is not clear to the uninformed, which could be countered by teaching the context of scientific discovery alongside it.

While doing outreach and decolonising in earlier education can be very valuable, however, some of us are a little bit skeptical when people suggest that since it is a frequent talking point of people who reduce structural issues to a “pipeline problem” (e.g. “how can we get black people/women to choose to do this subject?”). But this viewpoint can also come from frustration with the enormity of the task of decolonisation. How will people who don’t agree with the status quo achieve enough power to change things? And how do we dismantle something, without instituting another problematic framework in its place? Is it easier to dismantle something from within? Or is that becoming complicit in the process?

Other structural issues#

Other, perhaps even more insidious structural issues popped up in our discussions:

  • There’s this expectation to have “World leading papers” - and how the University measures research output quality is completely entrenched colonialism by language, networks, and popularity.

  • Language, both for reading/publishing papers and reading/writing code, everything must be done in English.

  • Knowledge in general is colonialised - how can universities/individuals in medium- and low-income countries afford access to extortionate journal fees etc?

What did you find the most and least persuasive argument in this paper?#

We are trying to include more critical questions in Data Ethics Club

Some of the feedback from our Data Ethics Club survey this summer was that the group would like the opportunity to be more critical of our reading. This is the first time we’ve included a question like “What did you find the most and least persuasive argument in this paper?”, but that’s not a reflection on this enjoyable paper! We hope to continue to include these kinds of questions to flex our critical thinking muscles and help us build our understanding of data science and ethics.

We commented on what was most and least persuasive to us, as people with an interest in data science and ethics, but not necessarily with either the academic expertise or the lived experience that might see another audience react differently. One of our main pluses and minuses weren’t really related to the arguments themselves, but the nature of the paper as a review-paper which can’t explain everything from scratch in 10 pages: this gives is an excuse to do more reading and a lot of jumping off points! It was mentioned that it would have been nice, however, to see an acknowledgement of what the paper did intend to focus on, and to (intentionally) leave out.

The article certainly didn’t pull any punches with its view of academia. Whether or not this is extreme cynicism or just unapologetically candid realism could be argued, but we certain found that aspect persuasive! We discovered a good opportunity for self-reflection in noticing that the arguments that we found ourselves agreeing with the most were the ones that we had personally experienced, e.g. the power of learning about the historical context of scientific discovery and how that changed our viewpoints. We wondered if others would have the same experience of confirmation bias, which could result in some people discounting the existence of scientific racism (although no one in our group held this view) and if there was anything the article could have done to anticipate that. Many of us also appreciated the explicit section on White Feminism, and as many of us are White women found it challenged us in a really useful way.

There was a lot of agreement on the point of the value of historical context, and how decolonising scientific curricula is a practical step that can be taken to help educate everyone. This course was recommended by the group as a starting point for learning more about what that might look like in your discipline.

One part of the article that some of us were uncomfortable with was:

“Black women, through years of training and enculturation in a white supremacist and colonialist system, are conditioned to internalize the status quo. They may thus be unable to describe and elucidate the systems that oppresses them
 They might lack the language to articulate it”

We considered this sentence to be making the point that Black women who endorse the current system as-is can’t be used to defend it, which we broadly agreed with; survivorship bias may also play a role here. However, we felt that the point was a little clumsily made, as it could be read as there only being one “correct” experience for Black women, or only some Black women who should be listened to, which we didn’t feel comfortable with (although we weren’t sure if this was the author’s viewpoint). This part of the authors’ argument is also unprovable (although, to be clear, not necessarily wrong), as any attempt by a Black Woman to oppose it can be deflected by stating that that person has been “assimilated” or by stating that they “haven’t educated themselves”.

The link between pseudo science and discrimination wasn’t super clear to some initially (“Are flat earthers inherently racist?”), but through discussion and the context of eugenics, it became clearer. Others found this point particularly persuasive, and to be a clear avenue for activism and convincing editors to do better. As the paper says:

“The system itself needs to be rethought — scholars should not, as a norm, need to form grassroots initiatives to instigate retractions and cleanup the literature. Rather, the onus should fall on those producing, editing, reviewing, and funding (pseudo) scientific work.” “Strict and clear peer review guidelines, for example, provide a means to filter racist pseudoscience out.”

The authors sum up a more nuanced point “Ultimately, decolonising a system needs to go hand-in-hand with decolonising oneself”, acknowledging individual agency, following their earlier persuasive point “centering individual people, as opposed to tackling systemic obstacles, is a myopic modus operandi and indeed part of the way the current hegemony maintains itself. Fundamental change is only possible by promoting work that dismantles structural inequalities and erodes systemic power asymmetries”.

What could you/your organisation do to decolonise your field?#

We’ve already discussed our favourite suggestions from the paper, but we had more to say.

The most important thing always has been and always will be to listen to people. Unfortunately, this is not always easily done with one or two voices dominating such conversations (one member of the group suggested a mute button for other people in zoom meetings). However, we can start with small discussions and form a ripple effect as we take the ideas out further, by holding/participating in workshops/discussions that are non-judgemental and learning environments like Privilege Cafe. Understanding your own privilege goes a long way to helping you check it. Once you’ve recognised it, then calling out your own mistakes (for individuals and institutions) is a start, and helps open up these conversations further.

Inspired by our previous discussion about “ideal roles”, one suggestion was for the work (of decolonialisation) to form an integral part of each job role. As the paper says, such work shouldn’t be done on top of a system that is itself inherently more-taxing for those that are oppressed: “these issues are difficult to acknowledge for those in power — they are seen as a sideshow, a political/politicised distraction rather than an essential element of good (computational) science.” This work should be valued in and of itself, not as a side hustle!

For those of who us who have begun to try, many of the things we try to do feel small and feel meaningless. Even small things get pushback, e.g. listing pronouns in email signatures. That doesn’t mean we should give up, but sometimes it’s helpful to vent!

Finally, collective action such as joining a Union (e.g. UCU and pay inequality) can have a positive impact.

Attendees#

Note: this is not a full list of attendees, just those that were comfortable sharing their names.

Name, Role, Affiliation, Where to find you, Emoji to describe your day

Structural Injustice and Individual Responsibility   UK National AI Strategy: Pillar 3 - Governing AI Effectively