MAA Hosts Ethics Workshop at New HQ

The Mathematical Association of America (MAA) recently hosted a workshop, Framing Mathematics as a Foundation for Ethical STEM, at its headquarters in Washington, DC. The workshop was run by Victor Piercey (VP), Catherine Buell (CB), and Rochelle Tractenberg (RT). Each brought their expertise in mathematics, law, and science in relation to the importance of teaching the value of ethical standards to our mathematical community. Below you’ll learn about their experience, teaching ethics in mathematics, and how you can implement ethical standards in your workspace or classroom.

VP: The first thing is to recognize that we are doing actions that have consequences. And recognizing that ethical decision-making is part of what we do and incorporating that in our work from step one. So what does that look like? The easiest cases are when there are applications to your work, it's less obvious when there aren't applications, but that doesn't mean it's not there. 

Everything from your choice to work on one problem instead of another could have ethical consequences. And maybe it's fine. Maybe there aren't any harms. It's all beneficial, but maybe there are. And having a little antenna up to think about that, even if it's just a pause, that's a starting point to say, what else could I be doing right? Who's funding my work? What's their agenda?

CB: When I talk to mathematicians about this, they recognize problems well. They see the societal problems that are going on with regard to trust and ethical issues. I think they see that mathematics has a role because mathematicians are notorious for recognizing [that] math is everywhere. 

Sometimes when I talk with mathematicians, we use the terms creators of mathematics and users of mathematics. This idea is that even if you are working in symmetric spaces and you aren't sure where the application is, there are still some ethical implications because you are part of a larger community that is a practice. 

What are the expectations for being part of that practice to ensure good decisions are made? Also, what are you doing when you interact with other people in the profession, potentially below you or a mentor, or with your students who will also go on and create or use mathematics? How do you want to ensure that when they go out into the world, they are mindful of the implications of both their work, how the public views their work, and how the practice, in general, is perceived by the public?

Those are important things to consider, no matter what you're studying. Scientists, engineers, biologists, and people in the medical field will move on. The idea that we don't have any ethical obligations or social responsibilities is neglectful of our place in the larger scheme. 

RT: I'm not a mathematician, and I have been engaged in ethical considerations in research for 30 years as a scientist after completing a PhD program in cognitive science. Research ethics paradigms tend to be concerned with how you interact with humans and that you're careful about the humans and, more recently, the humans’ data. But this model of how we train researchers is faulty because it assumes that: 

You don't know when undergraduates come into your class which ones are going into science. The dominant research ethics training paradigm is: ‘Don't train them all, and when they get to the point where they may encounter ethical challenges, that's when you train them.’ That's the current model. 

My work since 2009 has been to subvert that paradigm because it makes no sense. Not only are we selecting for 1% of our undergraduate population to become capable of at least contemplating the ethical implications of their work, but we're also assuming that none of the other people will ever surpass whatever barriers may prevent them from going into science and STEM, so we don't need to bother with them recognizing ethical challenges in research. 

In 2009 all institutions getting National Institutes of Health (NIH) funding were charged with documenting how they train new scientists in the responsible conduct of research. I had been working on a curriculum development and evaluation paradigm. My collaborator, an ethicist, and I built one of these curriculum structures for ethical reasoning. That effort was intended to make a developmental trajectory for ethical reasoning available, so undergraduates who engage in research or scientific thinking can get on this continuum at any point and start growing in this skill set, irrespective of whether they would ultimately go on to get federally-funded training in research. 

A unique feature of the curriculum paradigm is that learners can see their place on that trajectory and can choose to grow. They can begin to think beyond human and animal subjects’ care because ethical reasoning can apply to any kind of situation. Human and animal subject data is the main focus of efforts currently to train only those who get grant funding from NIH and the National Science Foundation (NSF). An important feature of this paradigm is the idea that plagiarism, falsification of data, and fraud are the only types of prosecutable misconduct. But the wider paradigm, for example, the National Academies, does not focus on only prosecutable misconduct. 

The National Academies have been focused on “disruptive research practices,” eliminating cherry-picking and making all science more transparent so that people are charged specifically not with just getting informed consent but also ensuring that their work is transparent and accessible and to limit misleading uses of their work. That model hasn't been adopted by the NIH or NSF, and the National Academies have no promulgation channels. 

As a statistician, I was the vice-chair and chair of the American Statistical Association’s Committee on Professional Ethics. It's a huge mission for such a small group composed of some professors but also people in the industry and in government who may be mentors and leaders but are not instructors. So, clarifying their role as potential channels for information about ethical practice - not just responsible research - makes it so that it's not only professors who have doctoral students that engage with the responsibility to train for ethical work. Similarly, the undergraduate instructors in our current project are important channels for this information about ethical math.

All of these features have been on my mind since 2009. So the opportunity to work with mathematicians has been fantastic because math and statistics are intimately related.

VP: Yes, I do. There's a lot of catching up the mathematical community needs to do. I feel like this is something a lot of us, myself included, when I was in graduate school, think of the work we do as being value-free, neutral, and not having consequences. 

Art is, for art's sake, in many ways, right? And we're finding that's just not true. 

It's hard to explain. It's still connected to the wider world and the systems, right? Lots of different systems, social systems, and a lot of that is getting weaponized now. Not all of it, but any work that's somehow connected to cryptography or data science and things like that can be weaponized against populations that are at risk. As a result, I think there's another reason that this is particularly important now, if not before, because of the way technology has grown. Whether we're talking about just machine learning or deep learning, artificial intelligence (AI) it's using deep mathematics. And there are mathematicians whose work is getting used, whether it's cryptography and used by the NSA to hack into cell phones, whether it's a web crawler that's scraping data of folks’ social media pages that they intend to be limited to their followers.

A number of technological developments use math, and some may involve math as collaborators in their development. I think it's critical. The work we do, while always powerful, is being made more powerful and more weaponized. And that's not what we want.

CB: You see the word [ethics] used more. You’re seeing blogs about ethics in mathematics from high up, including in the MAA. You also see the MAA talking about forming their own sort of committee to start thinking about what ethical practice means for them. You also see, I think the AMS Committee on the Profession hosts panels about ethics in mathematics. There have been special sessions. The Society for Industrial Applied Mathematics (SIAM) has also posted several things related to this topic. 

Many traditional mathematics organizations have started using the word ethics in their titles over the past couple of years. But there are a lot of people who have been talking about it who might not have used the word ethics. It's okay to recognize those things as well.

You are seeing ethics out there a little bit more, both formally and informally. For example, in any conversations you see people having about what it means to be in this profession and this practice, those are examples of ethical conversations. 

I think places where it's missing a little bit, as many of those are framed around what Victor just highlighted and sort of those bright spots where everyone can see the wounds of the use of mathematics. Sometimes the harder one to see is the ethics of the practice and the way it impacts other people.

Conversations about ethics in my classroom open doors for students who want to go into STEM because they're like, Oh yes, I want to be a good practitioner. It's also opening the doors to students who did not yet see themselves in STEM and say, Wow, these people think it's important to consider things outside their own silo. That opens the doors for more people to access the profession and the practice. 

I think that we should talk about that. 

I think that that is an ethical part of being a good mathematician. We are building trust in future people who will use and interact with mathematics. We are also framing the profession as one that does not point at others and say that they’re doing it wrong instead of recognizing that we all have a role in improving and that we should be mindful of what we do in our profession.

That's what I'd like to see more of; I know the fancy titles bring people there. Then it's the question of how we get there to stay there. That's why we started from the ground up instead of a pushdown approach.

RT: I'm a biostatistician, and when I first got into this space, my research ethics research partner, a priest (he has a PhD in ethics and also a PhD in microbiology), had been providing ethics advice on biomedical issues to the Vatican.

The Clustered, regularly interspaced short palindromic repeats (CRISPR) is a technology for editing genes and is one of these bright spots that Catherine was talking about. Technically you could have something that will be an organism when it grows up, clip or CRISPR something and make it have blue eyes and blonde hair, be super tall if you want, and then let it continue growing.

There are clear, obvious ethical implications for the use of CRISPR. And my colleague had been providing advice to the Vatican on issues like that. They're things that emerge. They're unpredictable because no one knew that CRISPR was coming except the lab producing it. Until it became available, ChatGPT was the same. No one knew that it was coming.

So, you have this catch-up problem when you only consider each ethical challenge as it comes up. An outdated idea is that we need a specific way of responding to each challenge. The curriculum development paradigm for ethical reasoning that we created in 2009 is the antithesis of that mindset. Training in ethical reasoning can be discipline-specific (like for math). Still, it is a general approach, promoting ethical practice as well as the ability to recognize new ethical issues and respond to them. 

That's been my focus. I see NSF, NIH and other funders, DOD, CDC, AHRHQ, all these different US federal funders have been creating specific funding streams for dealing with emerging ethical issues, which I think is completely wrong.

When you apply for a grant like that, you must articulate a particular ethical challenge, like what would happen if CRISPR got legs? Or what would happen if ChatGPT and CRISPR had a baby? I hope that (the funding approach) changes. Practitioners need to recognize the existence of a code of ethical practice but must also recognize that such a code will, naturally, change over time. So learning one code (particularly as an undergraduate) clearly isn't going to last you your whole career, but learning how to reason with whatever code or policy or law happens to be applicable in your space is a much much more productive way to get people thinking - and practicing - ethically.

In this paradigm for ethical reasoning, as Victor said, when you're presented with a case or a problem, you can determine, ‘Should I do that’ rather than just doing it and hoping it doesn't have a negative (unethical) outcome. Again, the funders are focused on graduate students who will be making new science or teas in the future or new CRISPR. This NSF-funded project is intended to promote a grassroots level of awareness that ethical mathematical practice is essential. 

The responsible conduct of the research paradigm is not trickling down from PhD programs to undergrads. So there are trends, but they aren't in the direction that I hope. I did another project in 2019 with the construct of disciplinary and professional stewardship. Stewardship requires ethical reasoning but includes much more - attention to promoting the integrity of your discipline or profession. As Catherine was mentioning, when we raise awareness amongst practitioners that they have some responsibility to their profession or the use of math in their profession, as a lawyer as a journalist, you have these ethical obligations to treat the tools you use respectfully and consider the impact on others of the tools you use.

A stewardly approach to your use of math as an engineer or physicist wouldn't go amiss. So this construct has multiple ways to engage with the thoughtful use of tools. That's what I'm hoping this work with math will do because it echoes the work we've been doing in statistics and in other domains to promote the stewardly use of math by, for example, - historians and neuroscientists. 

A stewardship model can help spread if your practice involves a lot of different disciplines. If you work in a multidisciplinary space, you can use an ethical reasoning perspective. However, the whole objective is to be transparent and respectful. 

We can ask undergraduates to think about the ethical aspects of their use of math. I wouldn’t say a wider appreciation for stewardship or ethical math-math/statistics is a trend. It's more anecdotal for me. Faculty are coming to me and saying, ‘I could do that in my class.’ So I hope this anecdotal experience and our grassroots approach become a trend because the top-down approach - training only PhD students and hoping somehow others also learn hoe to be ethical in their discipline - is hitting 1% of the college population. According to this paradigm, it will not get further than that.