Events & Calendars Faculty Focus

- I am Deborah Gewertz. Since 1977, I have taught anthropology at Amherst College, in courses such as The Anthropology of Food, the History of Anthropological Thought, and the one I think perhaps somewhat helped launch Carolyn on her way, the Anthropology of Gender. During this course, I spoke about Michel Foucault's famous book, "Discipline and Punish", in which he discusses various ways that bodies are controlled, especially through incarceration. I also spoke about the then-current highly distressing case of Susan Smith, who was convicted of murdering her two children, a case I believe Carolyn still finds relevant to her thinking about what she calls punitive motherhood. I should mention that this event is being recorded so that it can be shared with those unable to attend. I should also mention that during my conversation with Carolyn, the Q&A feature will be open for live submissions, and we do welcome your responses and questions. The Q&A box is next to the live stream feed. Today, Carolyn will be speaking about pregnancy behind bars. It's not something that most of us consider and even saying it out loud sounds a bit jarring, but Carolyn has spent much of her career working in this area and will be telling us why we should care about what's happening to pregnant people in these institutions. Before we dive into that issue, Carolyn, can you please tell us how you got into this work?

- Well, I, first of all, wanna thank everyone for joining today. And Deborah, what an absolute delight to be able to have this conversation with you. So how did I get into working with pregnant women behind bars? Well, it actually started when I was a first-year resident in OB-GYN. I was at the University of Pittsburgh. And it was late one night when I was on call, helping to deliver a baby. And by every other measure, this birth was like any other hospital birth. There were IV poles, there were nurses and doctors in the room, of course, a pregnant person pushing her baby out. But there was one thing that was different and this woman was shackled to her bed. And this was something I had never even conceived of could be possible. It was extremely jarring, as you said. And you know, when you're an intern, a first-year resident, mostly you're thinking about things like how can I make sure that this baby comes out safely? And how can I make sure I don't drop the baby? But in this particular instance, I was also wondering, what was I gonna do if there was an emergency and here she was shackled to the bed? And what is my own complicity in this? What am I doing here in this room? And it was a deeply troubling experience for me, nowhere near as troubling, I'm sure, as it was for the person who was giving birth. And in that moment in the delivery room, as all of these things were rushing through my head, I also really thought about all of the things I learned from reading Foucault at Amherst, the control of reproduction and sexuality, and incarceration. So it was really a very pivotal moment for me. And it made me realize, I'd never thought about the fact that there were women who were incarcerated in our country, let alone pregnant women. So it was really that experience that opened a Pandora's box for me, and that sent me along this path of launching the issues I was going to dedicate my career to.

- I find it gratifying, I must admit. Quite gratifying and impressive that you took my brief discussion of Foucault in class, such a long and deep way.

- Well, you know, there are a lot of ways that that class influenced my path as you rightly alluded to. This class, the Anthropology of Gender, really just transformed my worldview. And you know, it just... With this class, in addition to reading Foucault, you mentioned Susan Smith. And I obviously will never forget that paper because you had us respond to a headline from Time Magazine with her picture on it. And the headline was, how could she do it? And it was a challenge to our notions of the romanticized versions of motherhood that Western culture advances. And this has definitely played a role in my work and the way that we understand, and the way that the motherhood and reproduction of people who are incarcerated in our country is managed in a lot of punitive and repressive ways. And so that paper. But perhaps even more so, was the comment you wrote in the margins of that paper, which was, you should write more than prescriptions. You knew at that point that I was thinking about becoming a physician, and you saw something in my writing, and those words really, really encouraged me. And I think that was the first point of connection for us and how our mentoring relationship and friendship flourished from that. That class and that paper really had a long-lasting effect on my work.

- Well, thank you. In addition to being a practicing OB-GYN as you are, you are also a medical anthropologist. How did that come about? And how do those two fields intersect in your work?

- Well, you and the other anthropology professors at Amherst really instilled in me a deep connection to anthropological thought as my way of viewing the world. And I carried that with me when I went to medical school and when I started residency. And so after becoming interested and troubled by this area of women being pregnant behind bars I began doing a little bit of research on this topic when I was a resident in OB-GYN I decided to do a clinical fellowship in family planning. And when I was interviewing for fellowship positions, one of my interviewers at the University of California San Francisco, I was really trying to impress him 'cause that's where I wanted to go. He looked at my CV, which by the way, at that point was very bare. And he said, "Reproduction and incarcerated women. That's so interesting that that's gonna be your niche. There aren't really many OB-GYN who've looked into this." And I had no idea what my niche was gonna be. I was just trying to make it through the day as a resident. But I really wanted to go to UCSF, so I nodded my head vigorously and I said, "Yes, Dr. Darney, that's going to be my niche." And sure enough a year later, when I started my position at UCSF, I was encouraged to make connections at the San Francisco County jail. And so I did, and I started doing my fellowship research project looking at... A mentor of mine as well. He said, "Would you like to do some clinical work here at the jail?" And I thought, "Sure, that's my way of giving back. I'll do some volunteer clinical work a couple of times a month." Soon... I needed the tools of anthropology to help me sort through some of these contradictions. Like there were people who, the only place they accessed healthcare was in jail. I also learned that incarcerated people are the only people in our country with a constitutional right to healthcare. 1976 Supreme Court case Estelle v. Gamble. And the case made it all the way from Texas up to the Supreme Court. And it was based on this that the court said, incarcerated people, they don't have the means to just call up a healthcare provider or to present to a hospital, they rely on the prison. And so the courts determined that not to provide them healthcare was a violation of The Eighth Amendment, cruel and unusual punishment. So how unusual is it that someone becomes incarcerated, they get stripped of most of their rights, and yet they gain this right to healthcare guaranteed by the constitution? So that was another contradiction. And then for pregnant people, the only place many of them access prenatal care was when they were in jail. And so I began to learn about and experience as a caregiver, what these contradictions looked like. And I was also learning more about mass incarceration which at the time I was not as familiar as I am now with the phenomenon of mass incarceration and also, how it disproportionately affected women. And so as I was learning all of this, I sensed, as I mentioned, that there was something bigger going on. It was not only the narrative of warehousing individuals in prisons and jails and forgetting about them. Although surely that story of neglect is still relevant and part of the narrative. But there was something more about the nuance of the every day and that's why I needed anthropology. So I tapped into my networks and I think I spoke to over two dozen people, including of course you, Deborah about how to tap into my need for further anthropological analytics and whether or not I should go back to graduate school at the ripe old age of, well, somewhere in my early 30s but already being a practicing physician. And people like you, who've known me for a long time, who've known my connection to anthropology, encouraged me to pursue this. And so I did. And so I went back to graduate school and did my dissertation research at the San Francisco Jail. And the topic of that research became the source of the book "Jail Care".

- Well, you know a great deal about motherhood, reproduction and the country's punitive system. But could you tell us a little bit more anthropologically speaking, ethnographically speaking about what it's like to be a pregnant woman who is in jail?

- Well, there are many different ways that this plays out for people. And that has to do not only with the individuality people's lived experiences but also with the lack of oversight and standardization of conditions for pregnant folks behind bars. And so I'll get into some of the ethnographic detail in just a moment. But as background to that, the Estelle v. Gamble case I mentioned that established healthcare is a constitutional right for incarcerated individuals, it didn't come with any specificity. And the language of that court decision said that prisons and jails and other institutions of incarceration are required to address the "serious medical needs" of people who are inside. And if there are other healthcare providers on the line who are joining the conversation today you may wonder well, what do they consider a serious medical need? And the answer is, it really depends because there are no mandatory standards that every prison and jail has to follow. There is no mandatory system of oversight or accountability. And so it leaves things up to, a lot up to discretion of what healthcare, among other things, what healthcare services, including pregnancy, from prison to prison and jail to jail. And likewise, that also has to do with other programs and services they might provide for pregnant people. So let's take a deeper dive. So in my ethnographic research at the San Francisco Jail that I did for the book "Jail Care" and then subsequent research I've done at other sites has corroborated a lot of this. What I found was that many women really did rely on jail as their only source for prenatal care. So I'm gonna tell you all about a woman whom I call Evelyn, whom I came to know first as a physician at the San Francisco Jail providing OB-GYN care, and then later as I was doing the ethnographic research. And Evelyn was someone who had been in and out of jail many times, more than two dozen times in her adult life. And so when I came to know her when she was pregnant in jail it was not a new experience for her. It was not disruptive for her. And in fact, she had been in jail in previous pregnancies, as well as multiple times during this pregnancy that I came to know her. So again, it was not a disruptive event. Because jail was stuck, was folded into her life in such a chronic way, she found some familiarity in jail. She knew the deputies. She knew the other healthcare providers. She knew a lot of the other women who were incarcerated in the jail. And in fact, spent time with them not only in jail but also on the streets. And her life outside of jail was characterized by a lot of chaos. She, as a child, her parents died when she was, sorry. Her mother died when she was very young and her father and uncle molested her. This caused her a great deal, understandably, of trauma. And that trauma was never addressed. And so she turned to drugs as a way to self-medicate her pain and trauma. She never got the mental health care that she needed. And so for her, she became involved in the Juvenile Justice System and then as an adult in jail. So when she was pregnant and in and out of jail, this was a place where she could access prenatal care where she had a bed to sleep in, not a comfortable bed, but it was a bed. And she had parenting classes. She had yoga classes she could go to. It was San Francisco, after all. And there are a variety of services that she could get from jail, which is really troubling. It's jail. It's not a service provider. It's not a healthcare provider. And yet there were these services that she was accessing. Now, she, for the pregnancy that I came to know her she actually got released a week before she gave birth. Other women whom I came to know like a woman, I call Kima, she too, in and out of jail for much of her adult life. And in and out of jail during pregnancies. She did give birth while she was in custody. And for her, she could not have any family members with her in the room, the hospital room while she gave birth. And she did have the support of a doula who had supported her during pregnancy while she was in jail. But otherwise, she was fairly alone. And then what happened after she gave birth while she was in the hospital, was that, the sister who was supposed to be taking care of her baby, it turns out that those plans fell through. Child Welfare Services said the sister could not take care of the baby while Kima went back to jail because the sister had an open case with the Child Welfare System. So then the child was gonna go into foster care or if Kima could find another family member. But until that time, the Child Welfare Services worker put a police hold on the baby. And that meant that Kima couldn't be with her baby in her room. And mind you this is the only time she was going to have to bond with her baby in this immediate time period. And so she could only go see her baby if she was escorted with an armed guard to the nursery. And also Kima had to be wheeled while shackled to the wheelchair. So she could barely even hold her baby when she went to the nursery. And this experience was extremely painful for Kima. She then went back to jail where she tried to provide breast milk, to pump breast milk for her baby. That didn't last very long. But even these two snapshots, I'm not even sharing all of the details of their lives or their parenting or their being pregnant behind bars. These are very different in a lot of ways from other women I've spoken to at other jails and prisons and other court cases that I've read as well. Because there are also women who give birth at jails and prisons that are ill-equipped to handle pregnant women. And there are women who have gone into labor in their jail cells and given birth in their jail cells, places where women feel extremely isolated and have scant prenatal care. There are jails and prisons where if they need abortions and they can't get access to abortions because the institutions have decided that that's not a service they want to provide access to, even though the courts have affirmed that women retain their right to abortion. And then there are just the conditions of living every day in prison and jail. The mattresses they have are very thin and very difficult to sleep on even if you're not pregnant. But add to that the discomforts of pregnancy, add to that the changes to a pregnant person's nutritional-- some nausea and vomiting that some pregnant women experience. You can't choose when you eat when you're in prison or jail. There are scheduled mealtimes which are usually breakfast is at 04:30 or 05:00 in the morning, lunch at 11:30, and dinner at 04:30. And if you're pregnant and some of you who are on today's event maybe you have experienced pregnancy yourself know that sometimes you need to eat small frequent meals. That's just not a reality when you're incarcerated. So there are a lot of things about the daily living that are also challenging. And then if you do give birth in custody, as Kima was, you pretty much were separated from your newborn soon after birth with variable opportunities to actually see your baby. It looks like a lot of different things but for most people, it is a traumatic and difficult experience even at the same time that women like Kima and Evelyn might say that "At least I'm getting care in jail." but that's a problematic stance to even think about jail as a space where they can not access care, they should be accessing care. But to think of it as a place of relative familiarity and calm the way that women like Kima and Evelyn experienced the San Francisco jail, is really problematic. And I'll wrap up this comment about what it's like being pregnant behind bars and the troubling aspects of it with words quoting both of them. First Evelyn, she said to me, when she got rearrested, her third time on this particular pregnancy, she said, "Everyone says I got arrested, but I got rescued. I just didn't wanna get high no more. And even if it wasn't the best of medical care at least I was gonna get some medical care." And then Kima, who said to me on a number of occasions, both when she was pregnant and non-pregnant, she said, "You know, my worst day in jail is better than my best day on the street." And this is what I found in the research for "Jail Care". This troubling notion that jail was absolutely a space of deprivation and punishment, and at the same time, women like Kima and Evelyn felt a sense of relative safety, and stability, and care behind bars. That's a really uncomfortable finding when doing research about our institutions of mass incarceration.

- It's terribly upsetting. Pregnant women are in this circumstance, Carolyn. How many pregnant people are incarcerated in jails today? Well, that seems like it should be a really easy question to answer but it's actually not. So the Bureau of Justice Statistics which is a branch of the Department of Justice is congressionally mandated to collect certain demographic information about incarcerated people on an annual basis. How many people are in jails? How many people are in prisons? Sex, race, age, and reason that they're there and a few other things. And so we know that there are roughly two men behind bars on any given day in the United States, at least according to the latest statistics. But we don't know much more than that about the women in terms of their parenting status. A report from about 13 years ago from the Bureau of Justice Statistics estimated that two-thirds of incarcerated women are mothers and the primary caregivers to young children. And then in a 1999 report from the Department of Justice, they reported that six to 10% of incarcerated women were pregnant. But that's very old. And the means by which they gathered that data are not very clear at all. So until recently we actually had no idea how many pregnant people were behind bars and how many gave birth, had miscarriages, abortions, stillbirths, and other pregnancy outcomes. And so a few years ago as I was continuing on with my research about pregnancy and reproduction in incarcerated settings I began to question this citation that I replicated in a lot of my work the six to 10% of women are pregnant at any given time. And I began to realize that it really was not based on anything. And again, that we had no idea. And this, although the study I'm about to describe was not anthropological in its methods, it was epistemologically in questioning how we know what we know and what kind of information gets elided and why. So the fact that we didn't know how many pregnant people behind bars and how many gave birth it was, and is, deeply problematic because there are official channels for collecting the number of people who die in custody, but not how many give birth. And so I conducted a research study with my team at Johns Hopkins. We collected data for a year, from a sample of about 20, not about, 22 state prisons and all federal prisons and then a handful of jails, including the largest jails. So we collected data on this. And what we found was that if we extrapolated numbers from these sites nationally, that would amount to 3,000 pregnant individuals or 3,000 admissions of pregnant individuals to prisons every year in the U.S. and about 75,000 admissions of pregnant individuals to jails in the U.S. That is a lot of pregnant people who are going to need care, the full scope of care, who are going to if they give birth need new childbirth support and postpartum care. And we found that in our studies, prisons there were 750 births. And in the jails, mind you, there were only six jails in our study, there were about 130 births. So there are pregnant people behind bars and a lot of pregnant people who are admitted. I will say one of the most interesting findings was that there were very few abortions that happened and it was disproportionate to the number of miscarriages. And by disproportionate, I mean that miscarriages and abortions happen earlier in pregnancy. And it was not what you would expect based on national statistics and based on how far along people were in their pregnancies. And so I think what this study, in addition to just providing some data and the numbers, what it signals to me, that the fact that little old me and my research team at Johns Hopkins was doing the study and not a larger federal or a state agency. It signals to me that women who don't count don't get counted. And when women don't get counted, then they don't count. It just perpetuates their being neglected and ignored.

- In fact, it is a relatively small number of people in the grand scheme of things. And I guess a question that may be on some people's minds is why do you think people who don't work in prisons and jails, why should those folks be concerned about incarcerating, pregnant people, pregnant women?

- Well, the fact that it's a non-zero number to me, it signals that we should care. If there's one pregnant individual or if there's 75,000, this is a critical moment in an individual's life course. And it also has a ripple effect. This is, I mentioned that most incarcerated women in the U.S. are already mothers and primary caregivers to young children. So just incarcerating them in and of itself, has an impact on their families, on their children, and on their communities. And this is, although it's not my area of work to study the impact of having an incarcerated parent, there is quite a bit of work showing, not surprisingly that being a child and having a parent who's incarcerated has a lot of negative repercussions on your own mental health, and on that child's mental health, and on their educational attainment and their likelihood of becoming incarcerated themselves as adults. So there's really a ripple effect for incarcerating parents and mothers in particular, and then giving birth in custody or having other pregnancy outcomes in custody. Again, perpetuates this notion that it's all connected. Prisons and jails are not institutions that are over there. And I think we have seen that very starkly with the spread of coronavirus in institutions of incarceration, that if the virus doesn't know any boundaries all of the other social, political, economic, and healthcare phenomenon that affects people in the community are also relevant and in similar and in different ways for people behind bars. So it's really about all of our systems and how we value, and in many cases, devalue the reproduction of certain marginalized groups. And the way that I see my work is studying reproduction behind bars tells us so much about our social, economic, and racial dynamics and inequities in our society. Although I haven't talked about it in great detail, many of you in this conversation today, I'm sure are aware of the deep racial disproportionalities in our system of mass incarceration. And it is not a coincidence. There is a long legacy starting with slavery up to the, including the slave patrols the convict leasing system, the so-called war on drugs. Everything about our racialized foundations includes our system of mass incarceration. So motherhood behind bars is deeply connected to this because the women who are incarcerated in our country are disproportionately black women. And these are women whose pregnancies and whose motherhood, like Kima, who was a black woman, have also been intricately controlled by the Child Welfare System. So studying what happens, learning about what happens to women, to pregnant women, to mothers behind bars, to their healthcare to the ways their reproduction is marginalized and devalued in often very violent ways, this tells us everything about what is happening in our society. And I'll wrap up this comment by saying that an article that has influenced me by anthropologists Faye Ginsburg and Rayna Rapp is titled, I'm paraphrasing a little bit, but how reproduction dragged some central feminist concerns to the core of social theory. But just that studying reproduction tells us so much about our society as a whole.

- Thank you. We have questions, Carolyn and I will ask them. And they are from our audience. And the first is this, what impact has the COVID-19 pandemic had on pregnant people behind bars?

- Well, we know very little, specifically about pregnant people because as with pre-COVID times, they tend to be, not only marginalized but subsumed into everything else. My research team is in the midst of conducting a study of how pregnancy policies and services have changed right now because of the pandemic. And I can share some preliminary findings, also based on conversations I've had with colleagues who work behind bars and currently incarcerated women I've been able to be in touch with. Again, it's highly variable. There are some states and some counties who have prioritized pregnant people for early release from prison and jail, recognizing that they may be at higher risk for contracting the virus. And if they do, that can cause more serious effects on pregnant people. But for the pregnant folks who are still behind bars, it's an extremely, even more, isolating time because they don't know what's gonna happen to their babies. They don't know what's going to happen when they go into the hospital and deliver if they give birth in custody. Many labor and delivery units at hospitals have policies which have been evolving but especially at the beginning, there were policies that would separate moms from their newborns if the mom tested positive for COVID at delivery. For a pregnant person who's incarcerated the time in the hospital immediately after she gives birth is the only time she's gonna be able to spend with her baby. But there's also another finding that we've heard about, which is that, some prisons and jails are using solitary confinement or restrictive housing arrangements as medical isolation or quarantine, which should not happen. Solitary confinement is a behavioral punitive housing arrangement where people are in their cells for 23 hours a day and only get out for one hour a day. And there are already established adverse mental health impacts for people pregnant or non-pregnant of that housing arrangement. But when it's deployed as a health care measure for quarantine, that is extremely concerning. Then you add to that the fact that the national recommendations are that pregnant individuals should never be placed in this arrangement in part because of the mental health effects of it, but also because of concerns that they may not be able to expeditiously access medical care if they have acute needs like vaginal bleeding in pregnancy or contractions like they might be going into labor, but we're hearing that even for pregnant individuals, some facilities are using solitary confinement as a means of quarantine for COVID. But other than that, we don't know how many pregnant individuals behind bars have been affected. We do know the case that was widely publicized of a woman named Andrea Circle Bear, a native woman who was in federal custody and she died while giving birth postpartum on a ventilator in federal custody.

- A follow-up question by a different person wishes to know from you whether you have actually advocated for people to be released from prison, people with, for example, COVID to be released from prison.

- So I have been involved in some efforts, individual and as part of a working group to advocate for release of individuals, certain people who are at higher risk for serious illness from COVID, including pregnant people. So I have been involved in petitioning for individual cases and then I've also helped contribute to policy statements and national guidelines that promote depopulation as a means of controlling COVID-19 behind bars and in the community. And I have been that voice to try to advocate that pregnant people also be included as a potentially high-risk population.

- Thank you. Another one of my former students, Christine Hatch wants to know if black pregnant women's pregnancy healthcare outcomes versus white peers are they more or less equitable behind bars?

- Great question. As were all the others but you're tapping into something that I wish I knew and I wish that we could have studied in the study I described where we collected data. The short answer is, Christine, we don't know, and it's something that we would like to know. In the study I described, we collected aggregate data from each prison in jail. We were not able to collect individual-level data. It would have made it hard to convince the prisons and jails to voluntarily participate in this study if we ask them to tell us about each individual person. So we could not collect information about race. We do, of course, know. And I think your question was probably alluding to this that black women in the United States have a very high and disproportionately higher rate of mortality and morbidity in pregnancy. And that this is not a result of their race. it's a contribution of racism that is deeply embedded into our healthcare system, into our society, and to the resources that do not equitably support the healthy parenting reproduction and pregnancies of black women and other women of color. So I wish I knew, but I am certainly advocating for future studies. And in fact, future not research studies, but systematic data collection by federal and state agencies that collect data on pregnancy outcomes in prisons and jails and collected according to race. I will add that in the study I described, we did ask about maternal mortality and in our study, there were no deaths that occurred to the pregnant people during the time of our study.

- Thank you. Another follow-up question of sorts from Anna Correa Cosa, I hope I'm saying that correctly. Class of '09. And she wants to know what impact your own race and motherhood has had on how you approach to your area of focus.

- So it has changed over time because when I started this work, I was not a mother. And I became a mother as I was writing the book, as I was writing up the research. And then my race, my positionality plays an important role in that. I try to position myself appropriately and be forthcoming about the ways that my whiteness influences the ways that I see and ask questions, the limits to the information I can gather, and the analytics I can apply. It's something that I try to be forthcoming about with the women I work with who are disproportionately black women and other women of color, and in the ways that I write and position my findings and my analytics It's something, that it's not only about my research how does my whiteness affect the way I do my research, it's everything in life, as I think many of us have been and continue and especially in this moment are grappling with. So I can't say specifically my whiteness means that I see this finding in this way, but it does mean that I just try really hard in each scenario to ask myself, what does it mean that I'm a white woman asking these research questions? One way I do try to be humble and be aware of that is by seeking input and collaboration from directly impacted folks as I design research studies. And as I interpret some of the findings. And so I have colleagues, activists, and advocates and other researchers who are more directly impacted and they are women of color. So I see them as colleagues. But it's something I just I try to position myself realistically about. In terms of coming back now to the motherhood, I would say it has given me even more empathy to the difficulties of experiencing pregnancy behind bars and especially parenting and breastfeeding. Breastfeeding was an issue I thought about before becoming a mother, but I now see Kima's efforts to provide breast milk for her infant were extraordinary. She would sit in her jail cell and pump every few hours with a handheld pump, which at the time, I didn't really even know what that entailed. And she would do this. And she only did it for about 10 days, but she did. I remember, I sometimes actually helped. Her baby was still at the hospital, and so I would sometimes go from the jail and transport the milk after it had been labeled and everything on ice. I would ride my bike across San Francisco to the hospital. I now see what an effort that was to not have the direct feedback from your own baby to then pump breast milk. She barely saw her baby. Although she did, she had a few supervised parent visits when she was in jail. So I would say that being a mother gives me even more empathy about how difficult it is to be pregnant and postpartum and incarcerated. I remember also Evelyn she gave birth, as I mentioned about a week after she got released from jail. And although she was in a drug treatment program that was specifically designed for pregnant women or for parenting women with substance use disorders who were in treatment, so that they could bring their babies, live with their babies at this treatment program. She was not yet allowed to have her baby with her because she had not proven herself adequately for the judge and the social worker. So I watched her in the first days and weeks that she was postpartum, trying to petition for her right to see her baby. I would drive her to supervised visits in this tiny room across town where she got to see her baby for two hours a day, three times a week. And she would be trying to breastfeed her baby while Child Welfare Services worker sat in the corner with a laptop taking notes on every way she was holding her baby. And at the time, I thought that was hard and troubling, but I think I have a deeper sense of empathy of how difficult it is. And then bringing the race factor back to it, being a white mother, how privileged and easy my motherhood is as opposed for Evelyn and Kima, who from the start because they were black women within the criminal legal system they were already in a way, their motherhood was already written off and devalued, and they started out having to prove their worth as mothers, rather than for me as a white woman, just being accepted that I could express my motherhood in various normative ways.

- Theodore Hertzberg, Class of '04 tells us that he is prosecuting a pregnant woman in a violent armed bank robbery. And he wants to know whether you advocate an alternative to incarceration.

- So in general, for pregnant people who are incarcerated I do advocate for alternatives to incarceration when it's appropriate. As a healthcare provider and anthropologist, I'm not as deeply attuned to the legal scenarios of when it may or may not be absolutely appropriate. But what I do know is that most pregnant individuals have not been convicted of violent charges. And even if they have, this is less getting to the alternatives for incarceration but regardless, pregnant individuals must have access to full-scope comprehensive quality reproductive pregnancy care behind bars, which is not consistently the case. But in terms of advocating for alternatives to incarceration, I do advocate for it for most women. I would say I haven't been involved in enough cases of people who are charged with violent and convicted of violent offenses to be able to comment on when that level of charge is, or is not appropriate. But I will also say that for, especially in getting back to the case of coronavirus, we've seen, for example, in the state of North Carolina, where they before the pandemic had nearly 50 pregnant people in custody on any given day, they very early on instituted policies that released pregnant individuals early, prioritized them as a population for release. And I think after a short period of time only then have less than 10 pregnant individuals. And so there was still some people behind bars, pregnant individuals behind bars but they showed how very quickly it was possible and safe to release these individuals. And I'll close out this comment by saying, in terms of advocating for alternatives to incarceration for pregnant individuals, whether they're charges are violent charges or nonviolent charges, in some ways this is part of a bigger conversation of our entire carceral system, and whether we should be relying on incarceration at all as a means of addressing social, political, and economic issues. So I think that's a larger conversation about the role of our carceral system but to note that for pregnant and parenting individuals, most of the women who are behind bars are there because of things that don't necessarily relate to them being threats to public safety or to other people.

- Carolyn, we have time for one more question, I've just been told. And we have many, many questions. And I will send you these questions so you can contemplate them at your leisure. But let's take this one from Diego Duckenfield Lopez, Class of '24. He wants to know, how do the reports of hysterectomies in ICE detention camps fit into the history of reproductive rights in jails?

- It is entirely consistent with what we know about what happens to reproductive rights and healthcare behind bars. And I'm actually going to expand that to say that it's actually more broadly an issue of reproductive justice. This is a framework that I take a lot of teaching from. And many of you may be familiar with it, for those who are not, reproductive justice is a framework for action, a theoretical framework as well, that emerged from and centers the experiences of black women and other people whose reproduction has been marginalized and devalued. And reproductive justice emerged in part as a response to the reproductive rights movement. And I promise I'm gonna answer your question but this is important context for how I think about this issue. It emerged as a response to the limitations of the reproductive rights movement, which was in many ways, still is focused on an individual sense of rights and was focused on the concerns, at least initially in the '60s and '70s and beyond of white women, and whether or not they could have access to contraception or abortion so that they could stop having babies when they didn't wanna have babies. And the black women who founded the reproductive justice movement and framework saw that is extremely limiting. It was Civil Rights individual notion and not a Human Rights framework and didn't incorporate a broader sense of structural inequities and the material resources that were needed to enact one's ability to have healthy reproduction in broad sense. And so there are three core rights to reproductive justice, the right not to have children, so the right to access contraception and abortion, but also the right to have children, and the right to parent in dignity and in safe environments. And the black women who founded the movement recognized the long history of the ways that slavery that the genocide and controlling of the reproduction of native American populations, eugenics for sterilization of immigrant populations of the "feeble-minded", the ways that a long legacy and a contemporary reality of controlling the reproduction of marginalized groups, whether that's preventing them from accessing contraception and abortion, or whether it's suppressing their reproduction. Reproductive Justice recognizes that these are all part of the same conversation and all part of the same forces of reproductive oppression. And so when this report of alleged excessive and coersive hysterectomy of women in immigration detention came out a few weeks ago, it was both shocking and not surprising. Because it is part of a contemporary reality and the historical legacy that devalues the abilities of all people to reproduce in the ways that they want to. And hysterectomy is precluding a woman's ability to reproduce in the future because it removes her uterus. And this was also in, many people in this conversation may be aware of this report of the ICE hysterectomies, but you may not be aware that very recently, there were also, not just reports but documented cases of coercive sterilizations of women in prison, in the California Prison System. And the state of California has since banned sterilization procedures for incarcerated people. But this is happening for people in immigration detention. And it's happening for people who are incarcerated. California was just a state where investigative journalists did this inquiry. I wouldn't be surprised if it's happening in other States as well. So Diego, this was a really relevant question and I'm glad you brought it up because it's important that we think of these issues of reproduction behind bars as not merely a matter of reproductive rights or access to reproductive healthcare, but thinking of it in the very broad and structural sense of the reproductive justice framework that recognizes the ways that our society devalues and doesn't necessarily equitably promote the reproduction of all individuals.

- Well Carolyn, I imagine that we could continue. And I see the questions continuing to come in all evening. But we will have to call it a day. We're running out of time. And I would like to invite the audience. I don't know how they're gonna thank you exactly given the fact that we're on Zoom, but I'm sure they would all love to thank you for a wonderfully engaging instructive, important conversation with us. So thank you. Thank you on behalf of everyone.

- Thank you.

- I am to tell you that our next talk will take place on October 26th, when Sarah Bloom Raskin, Class of 1983 and Ron Lieber, Class of 1993, provide thoughts on financial resiliency and how Americans move from recession to rebound. Everyone, thank you. Thank you. And have a good night.

- Good afternoon, welcome to this Amherst conversation, "Will He Go?: Trump and the Looming Election Meltdown in 2020." Thank you so much for joining us. I'm Catherine Epstein, provost and dean of the faculty and professor of history at Amherst. It's a great honor for me to introduce the participants in this Amherst Conversation today. Let me start you with Laura Jarrett, class of 2007. Laura is the anchor of CNN's Early Start with Christine Romans. Since joining CNN in 2016, Laura has covered many significant legal stories, including the travel ban and migrant family separations at the Southwest border, the firing of Attorney General, Jeff Sessions and FBI Deputy Director, Andrew McCabe, and high stakes battles between the justice department and Capitol Hill. Laura had a lead role in delivering the breaking news of special counsel, Robert Mueller's appointment to oversee the Russia investigation in May 2017, as well as the conclusion of his work in March 2019. She also served as the lead reporter for CNN in the corruption trial of Senator Bob Menendez, and she covered the justice department Inspector General's report on the FBI's investigation into Hillary Clinton's handling of classified information. Prior to CNN, Jarret worked as an attorney in Chicago. She graduated from Harvard Law School in 2010. And while in private practice, Laura focused on defending companies and individuals in government investigations brought by the justice department and securities and exchange commission. Laura, thank you so much for joining us today, it's wonderful to have you back in Amherst. Now, Laura will be in conversation with Lawrence Douglas, the James J. Grosfeld Professor in the Department of Law, Jurisprudence and Social Thought. Lawrence is the author of seven books, including "The Right Wrong Man: John Demjanjuk and the Last Great Nazi War Crimes Trial" published in 2016 and a "New York Times" editor's choice. Lawrence has also published two novels, "The Catastrophist" in 2007, a Kirkus best books of the year and "The Vices", a finalist for the National Jewish Book prize. His commentary and essays have appeared in a host of high profile journals, including "The Wall Street Journal", "The New York Times", "The Washington Post" and "The Los Angeles Times". He is also a regular contributor to "The Times Literary Supplement" and "The Guardian". Today, Lawrence will discuss his most recent book, "Will He Go?" published in late May. "Will He Go?" explores a really intriguing question. What happens if the 2020 election results do not favor Trump and he refuses to concede? "Will He Go?" has received widespread praise and has attracted national and international attention with feature articles in "The New Yorker", "The Washington Post", "The Atlantic" and "The New York Times". The book has also been featured in leading publications in the United Kingdom, Germany, France, Spain, Ireland, and many other countries. Lawrence is donating the earnings from the sale of his book to The Hunger Project, a COVID relief charity. Lawrence, thank you so much for talking about your book today. I want to remind the audience that this is being recorded so that those who cannot come today will be able to see this conversation and now I'm going to turn it over to Laura.

- Dean Epstein, thank you so much for that introduction. I'm just so thrilled to be here this afternoon with one of my all-time favorite professors, professor Lawrence Douglas, to talk about his important book at such a historic time for our country. When I graduated from Amherst College with a major in law, jurisprudence and social thought, sometimes people would ask me afterward, "Did you always know that you wanted to go to law school and practice law?" And I would say absolutely not to both of those questions, but LJST classes and professors do have a unique ability to make legal issues matter, they help us identify weaknesses in our system and to light a fire under us to do something about those weaknesses. And those are all qualities on full display in Professor Douglas' latest fascinating book, "Will He Go?" There are now 54 days until the November election and I find it hard to imagine a timelier topic than the ones so deeply interrogated by Professor Douglas. So everyone knows what to expect this afternoon, I thought we'd start our discussion today, exploring how this book was born, what I see as its central premise, that Trump will not go quietly into the night should he lose the November election, then we'll wade into a little bit of the history on when things have turned ugly in the past and then finally, I thought we would chat a little bit about some of the possible scenarios of what Professor Douglas predicts will happen on November 3rd or at least after November 3rd and then finally, of course, we will leave plenty of time for all of your questions there in our virtual audience. And I'm told that you can locate the Q and A box next to your live stream feed, so feel free to shoot us lots of good questions and we will get to them at the end. So let's get to it, Professor Douglas, I thought we'd start by hearing from you about why you decided to write this book, given all the ways that the president has flouted so many of our constitutional and legal norms and ethics. I'm curious how you decided to focus on this potential disaster instead of focusing on all of the existing arguable breakdowns in the rule of law.

- Well, Laura, before I even start answering that question, let me just participate in the mutual lovefest because it's a really an honor for me to have you asking these questions and you remain in my mind as one of the great students to pass through Amherst during my time there, and it, there's nothing more gratifying for a professor to see a former student go on to do great things. So it's a real delight to have this conversation with you.

- Thank you.

- You know, I started actually kind of thinking about this book all the way back at the time that Trump had been inaugurated. Shortly after the inauguration, I started writing opinion pieces for "The Guardian". And I think the very first opinion piece that I wrote from "The Guardian" was about Trump's claim that you no doubt recall that he, in fact, would have also won the popular vote in 2016, had it not been for the presence of three to five million phantom illegal voters. And at the time a lot of people were thinking of this is just an example of Trump's extravagant narcissism, he needs always for his to be the biggest and the best, his inaugural crowds have to be larger than President Obama's inaugural crowds. And in this early piece, I kind of really ask myself, you know, what are the politics behind a claim like that? And I really was kind of curious to kind of start thinking about, well, what would happen? What would happen if actually a sitting president of the United States, if a sitting president of the United States actually claimed that he had been re-elected but for some kind of corruption or some kind of hoax in the electoral process. And so it really goes all the way back to 2017 that I started kind of thinking about this issue. And I really should probably also mention that when I started writing the book, I thought of it really more almost as a thought experiment because I wasn't necessarily assuming that Trump would lose. I was actually almost working from the assumption that he looked like a pretty strong candidate at the time or pretty strong a candidate for re-election, and I was just kind of curious and asking, you know, well, assuming that you did have an incoming president who was challenging his defeat, how well is our constitutional and our federal system equipped to deal with that type of challenge? And I think, I guess one of the big conclusions of the book is it's not particularly well equipped at all.

- Right, and I want to spend a fair amount of time on that, but before we get to the system itself, I wanna take it back to Trump because think you set up a premise here that I want to explore. And as I see it, the premise is that there's an inevitability of Trump's refusal to accept defeat. Should he not be the victor? And so you write this quote, "While his defeat is far from certain, what is not uncertain is how Donald Trump would react to electoral defeat, especially a narrow one, he will reject the result." It's a pretty definitive statement. And so I wonder, what is your evidence that this man actually wants to be president? And I put it to you that way because we just look at all of the self-defeating behavior, there's such a long litany of examples. And only the latest is the fact that he sat down with Bob Woodward, one of the most famous journalists in the world, perhaps best known for taking down a president. And he gave an on-the-record interview saying a whole bunch of things that for any other president would be simply the kiss of death. I wonder if that is not an indication, this is not somebody who wants to win.

- Right, okay. Maybe they're two separate questions. One is the question about, you know, how can I confidently predict, you know, that he is going to refuse to concede? And then next, maybe the second question is, especially in the light of the fact that maybe he doesn't even want to be president. Now, maybe I'll start with that second question first because I remember even back in 2016, and I remember back in 2016, that all sorts of people were writing things like, "Wow, imagine how dangerous Trump would be If he weren't such a transparently self-destructive person. Imagine he could actually win the presidency of the United States." You had all these journalists writing things like that. I remember Frank Bruni in "The New York Times" writing about, "Well, thank God we have someone who really doesn't even want to be president, he's so self-destructive that he, all he does is shoot himself in the foot. God, we would really be in a dire situation if he weren't so transparently self-destructive." Well, so much for those kinds of responses to Trump's presidency. And in fact, you know, we could have a whole long discussion about the simple fact that Trump has engaged in acts that you would think are incredibly self-destructive, and yet he ultimately hasn't had to pay the political price for them that one would predict that he would have had to pay a political price for, and that itself, it's a very concerning aspect of our contemporary political reality. Then the question is this claim that I make, that he's not going to concede. In the book, I do try to make a distinction between conceding on the one hand and submitting to defeat on the other. And conceding, I really kind of think of as a normative act in which a candidate recognizes that they've lost legitimately, that is that they've waged the good fight, and they recognize that they have been defeated and they kind of recognize the legitimacy of that defeat. I find it virtually impossible to imagine Trump doing that. And I think that we have a strong enough record of his past actions and behavior to support that conclusion. That is if we go all the way back to the Iowa caucus in 2016, you know, as you recall, Laura, he lost the Iowa caucus to Senator Ted Cruz of Texas. And the very next day he tweets saying, "I didn't lose, it was a hoax, the victory was stolen from me, Cruz cheated, and the whole Iowa caucus should be a run all over again." And we've seen this as kind of a consistent pattern on his part. I mean, if you really weren't interested in being re-elected, why is it that he is spending so much time telling the American people that our electoral process is not reliable? I mean, he really has created this kind of heads I win, tails, you lose scenario, in which the only way the electoral process can demonstrate his legitimacy is by him winning. And if he loses, it simply becomes defacto evidence of what he's been saying all along, that the election will be the greatest corrupt election in American history, the greatest hoax in American history. And, you know, it's an incredibly damaging thing to our democracy to have a sitting president telling the American people that the process is going to be corrupt if the margin of victory turns on these mail-in ballots, and I can't see why he would be spending so much time messaging in that way if he weren't really interested in being re-elected. So I think this is a guy who wants to be re-elected and is not going to concede defeat. Now, that said, it is possible that he could submit to defeat. And submission, I think, simply is a recognition on the part of a person seeking higher office that additional fighting the result is futile. And I think it is possible that if he loses very decisively, he will submit to defeat, but I cannot imagine, Laura, him conceding to defeat, I just don't see it in his political DNA.

- So to flush that out, what if one of the possible reasons that he's tweeting nonstop about a so-called rigged election and warning his supporters, you know, don't let the Democrats steal this away. What if one of the reasons he's doing that is to lay the groundwork for a submission of defeat, if you will, so that it is just less embarrassing because I think one of the ways that you sort of lay this out, that's really interesting is to sort of document all the ways in which Trump is so protective of his brand. And the part of the reason he doesn't want to lose here is not only the fact that he could be subject to prosecution by the SDNY when he's out of office, and there's all kinds of interesting issues with that, but that if he loses it sort of tarnishes this brand and bravado that he's worked so hard to cultivate with his allies.

- Yeah. Yes, that's exactly right. And in fact, you know, in a certain sense, this goes back to the, your earlier question, Laura, about whether he really wants to be president, because I think we could have a very interesting discussion about Trump's attachment to the presidency because he doesn't really seem to want to be president in order to exercise power in a kind of a conventional political sense. It seems more as if the presidency is a device that permits him to be the center of attention and being in the center of attention and brand protection and promotion, seems to be his chief political aim, rather than, again, exercising political power in any kind of conventional sense. But that said, I do think brand protection is absolutely critical to him, protection and promotion. And in that sense, a defeat is simply not an option because he cannot be presented as a loser. So again, he would rather kind of run-down American constitutional democracy than let himself appear to his base as someone who has been legitimately defeated. And of course, again, that's, it's a very, very unfortunate thing that Trump doesn't see any kind of light between his political fortunes and the nation itself.

- Well, one of the ways he certainly stays the center of attention is by consistently talking about a rigged election, I think it puts the press on edge and it obviously makes democrats on Capitol Hill flip out mostly because I think that there's just this fear of the unknown, and what exactly is going to happen. And I want to get more into that in a little bit, but I also hope that we could touch on just a little bit one of the things that I think is crystal clear from reading this great book, which is the way that our constitutional system just simply isn't prepared for a president like that, like Trump, and we've seen this play out in a lot of different ways, but on this point, you write the following, which I think really illuminates it, you say, quote, "Our constitution does not secure the peaceful transition of power, but rather presupposes it." So in that way, we've obviously faced electoral nightmare scenarios before namely in 1867, '76, I should say, and with Bush v. Gore in 2000. But I get the sense from reading this book and the way that you sort of make your argument here is that the situation here could be far more dire and so help the audience understand why.

- Yes, and I think that the point that you make is a very important one. So even if we go back to, since I think most of the people tuning in right now don't recall the election of 1876, why don't we look to the election of 2000, the election between Al Gore and George W. Bush. And as you recall in that particular election, the outcome was uncertain because the state in which the outcome, in which the electoral count outcome would turn on was Florida, and in Florida, there was this improbably narrow margin separating Trump from, separating Gore from Bush. And 35 days after the nation had gone to the polls, that's when the Supreme Court stepped in, in this Bush v. Gore decision, essentially stopping the Florida recount with George W. Bush clinging to a 537 vote lead. Now that's, I mean, obviously a ridiculously close result, and a lot of people, I think, credit the Supreme Court with ending that electoral dispute that is even if they believe that the Supreme Court decision was transparently partisan, and even if they think it was a poorly-reasoned decision, I think a lot of people say, "Look, well, we needed some kind of necessary closure. We couldn't let this thing escalate into a real electoral crisis and so the Supreme Court had to step in and end things." I think that that is a mistaken way of thinking about what happened in 2000. I do not think it was the Supreme Court that brought closure to that electoral dispute, I think it was Al Gore who brought closure to that electoral dispute because on the day after the Supreme Court rendered its decision, Gore conceded and he conceded very graciously. And what he did is he basically put the nation's political interests ahead of his own political fortunes and in saying that, you know, I'm not necessarily saying that Al Gore is a true American hero or a wonderful human being, now, he very well might have those characteristics, but what Gore clearly had is he had internalized the norms of constitutional governance, and he was interested in putting the interest of the nation ahead of his own political fortune. And I think if you had slotted if you'd slotted Donald Trump into the year 2000, there's no doubt that we would have had a real electoral meltdown because I find it impossible to imagine Trump conceding under the circumstances in which Gore did. And we should bear in mind that there were many people within the democratic party, other Democratic lawmakers, and people within Gore's legal team who really were urging him to push the fight and so, and he just chose not to. And again, this goes back to the point that you're making, Laura, which is the importance of norms. You know, laws can just, they can sanction us, they can deal with certain forms of truly egregious behavior, but really for a constitutional democracy to work, there's just a deep fabric of norms that need to be respected in order for the, really this system to function. And I think we've seen Trump's not simply pushing against those norms, but basically rejecting them outright and that's why I think we could very well be in for very, very destabilizing months following November 3rd.

- So given how tight things were in Florida in 2000, I think that along with obviously the situation in 2016, where Hillary Clinton wins the popular vote, but ultimately loses the election has led some commentators to suggest that Trump needs to lose the popular vote basically in a landslide and also lose the Electoral College vote handily in order to walk away, maybe not concede, but submit to defeat. And I wonder if what's the scenario in which you envision that happening? In other words, how badly does he have to lose do you think for that to happen? And is that even a conversation, I guess worth having, because this race is actually going to be pretty close.

- Right, well, again, maybe we're expressing a kind of wishful thinking rather than entirely predictive thinking, but I think you are absolutely right to suggest that in order to avoid kind of Trump playing constitutional brinkmanship, I think he does need to be beaten decisively and it raises the question of what we mean by decisively and obviously beating him decisively in the popular vote is not going to be sufficient. We saw that Hillary Clinton beat Trump quite decisively in the popular vote and given the anachronistic, and I would say really kind of dysfunctional method by which we elect a president, the popular vote, the loser can obviously win the Electoral College. And I would add one more thing to what you said, Laura, which is I think it would be important for Trump not simply to lose decisively in the Electoral College vote, but I think it's important for him to lose decisively in the swing states and for that defeat in the swing states to be apparent relatively quickly after November 3rd, that is given the fact that we're going to be voting in a time of pandemic, it's very likely that we will not necessarily know who has been elected president on November 3rd itself, but I think it would be important for the election not to drag out for weeks. I think it would be very important for us to know early on that Trump has lost the swing states as the count continues in these swing states, it's simply in a sense confirming the trends that we've already seen, that he's lost these swing states decisively and I think that would save us from a kind of electoral meltdown in this season, so that's the aspiration. Now, of course, the question that you again raise is whether the reality is going to match our hopes.

- Well, so speaking, I guess, of what happens in the swing states, this idea of a so-called blue shift, where there are obviously a number of outstanding mail-in ballots because of COVID or provisional ballots, just for whatever reason leads us not to have a result where Trump is defeated overwhelmingly in those swing states, but rather, it's either kind of on the margin or it's, you know, tight. And then over a course of, let's call it even just a week, it starts to swing to Biden as more of those ballots come in, the absentee ballots because they tend to be for democrats. It's one of I think what I might call doomsday scenarios that you offer in this book. The other one, just so that people know and hopefully gives them a reason to want to read this if they haven't already is this idea of dueling certificates, which has happened in our history before as to which candidate a state goes for, talk just a little bit about that so people realize that this can actually happen.

- Okay, so maybe, let me start with the first thing with the blue shift because again, this is something where I think that this is one of the things I was speculating about, but I hate to say this, but I do think that this is, it's starting to look almost like, a train wreck in slow motion. It seems like it is not unlikely that this is exactly what's going to happen. And the reason I say that is let's just kind of connect a couple of dots. One dot is the fact that we know that Trump has been very aggressively trying to insist that mail-in ballots are unreliable, that they're going to be corrupted. And one of the questions you can answer is why this attack on mail-in ballots per se. And I think a poll that was taken, or at least the results were, I think announced on August 26th, is pretty revealing in supplying an answer. This poll indicated that about 20% of Trump supporters plan on relying on mail-in ballots, about 60% of Biden voters answer this poll saying they plan on using mail-in ballots and so this is kind of part of this politics that of basically trying to discredit votes that are going to swing in Biden's favor. I would connect that to one other point that Trump made which I thought was a very telling tweet. And this was a tweet that he issued on July 30th and as your recall, Laura, that was the day that he floated this kind of crazy idea of delaying the election. And a lot of media attention was glommed onto that and pointed out, you know, what a crazy idea, he can't even do it. Congress sets the date of the election, so what's Trump be wise, even floating this kind of outlandish idea. That same day, later that day, he actually issued a second tweet that got very little attention, it was very short. And one of the things he said is, he said, "We need to go with the November 3rd results, can't wait days, weeks, months, years later to get the final result." And I thought that was incredibly telling, and I think if you put those two things together, his attack on mail-in ballots and his insistence on going with November 3rd results, we see kind of the architecture of how he could try to really gum up the works this coming fall. Because I think what he will try to do is whatever lead he enjoys on November 3rd, and I think his best chance of re-election is to lead on November 3rd because again, that's before all these mail-in ballots, that are going to break for Biden have been counted. And I think what he'll try to do is he will try to leverage whatever lead he enjoys in November 3rd as an indication that he has been re-elected. And then, if it is the case that that lead seems to evaporate, particularly in swing states as a result of the late count of mail-in ballots, well then he's going to kind of trot out the argument that well look, this is what I've been saying all along. I predicted that the mail-in ballots would be corrupted, they have been corrupted and we can't accept those results. And the way in which that kind of claim could really gain a lot of traction is again, if the margin of Biden's victory turns on the late count, the late count of mail-in ballots of the swing states and on narrow margins, then you are really in for, I think, a very, very ugly time, because I do think that Trump's attack on the mail-in ballots and the insistence that we go with November 3rd really could gain traction, that has gained traction both with his megaphones in the right-wing media, who will be reliably repeating and amplifying his claims. We could also imagine, of course, we could add to this mix, disinformation campaigns, either domestic or international, it can be from Russia, also trying to insist that the mail-in ballots have been corrupted. So all of that, I think, really becomes a recipe for a very, very concerning fall.

- Well, and certainly one of the things is, as you so nicely explained in the book is, you know, our system is set up with states, everyone does it so differently, and so for some mail-in ballots where this, you know, the states allow it, they don't even begin counting them until November 3rd, whereas in other states, they do have them counted, we don't know the results, but they don't even begin the process and so it could easily drag this thing out. I know I am prepared for a lot of late nights that week and frankly, for the rest of the month of November. We have a number of questions already in the queue for you.

- I'm sorry, Laura, let me just comment on that because I think that's also a really important point that you bring up that many of us are aware of this, but you know, we don't have a federalized election system. Everything is basically kind of left to states and in fact, lots of things are left to localities within states. We have well over 8,000 separate election precincts in the United States, which can have their own little separate ballots, they can have their separate technologies. And one of the things that I think is worrisome is as you pointed out, there's some states that as soon as they get their mail-in ballots they start counting them, so for example, Colorado, as soon as they arrive, Colorado election officials start counting them. If we look at the three swing states that really could determine this election, Pennsylvania, Wisconsin, and Michigan, the same exact swing states that basically handed Trump the election in 2016 recall by a margin of only about 70,000 votes in total between those three states, that's what gave him the presidency. All three of those swing states unfortunately have state laws, which prohibit the counting of mail-in ballots until election day itself and so, again, that just becomes a recipe for the possible delays. And we can also rest assured that Trump is going be sending out his legal teams to challenge the submission a lot of these mail-in ballots, because there will be legitimate legal issues to raise because again, states have various laws about when they have to be received by, in order for them to be qualified and counted. And to the degree to which Trump's legal teams can litigate the submission, can litigate these mail-in ballots, arguing that they need to be disqualified because submitted in an untimely fashion and the degree to which they can drag out the counting process in the swing states, it all breaks to Trump's favor.

- Well, I think it's one of the reasons that so many people have found it troubling that the Attorney General, Bill Barr has been so outspoken against mail-in ballots as well. He's obviously not technically the president's personal attorney, but having the weight of the justice department marshaling in favor against legitimate mail-in voting has not gone unnoticed, I should say. I don't want to hog up too much time, even though I could talk to you for hours about this. I do just obviously given so many alumni and students hopefully who are joining us today, before I turn it over to the questions from the audience, I would love to just know how did teaching at Amherst prepare you to write this type of book?

- Yeah, that's an interesting question, Laura. Maybe I have to answer it this way. So as I was writing the book, I participated in a number of workshops that were basically conducted largely by law professors, both on the left and the right who were concerned about the possibilities of this election going sideways. And a lot of these are law professors, they're really experts in, let's say federal electoral law in ways that I frankly am not. But one thing that I did notice is perhaps as a result of teaching in law schools and you and I are both graduates of law school, so I think we, this experience is familiar to us, they tended to get very much in the weeds and really not to see the broad picture. And I think one of the things about teaching at Amherst is it really kind of encourages you to think broadly and to make kind of broad conceptual connections. And in some of the courses I've taught over the years, I mean, I continue to teach a course basically on the Supreme Court and judicial review, I teach a first-year seminar on the rule of law. A few years back, I co-taught a colloquium with my colleague, Austin Sarat, on the meaning of catastrophe and a lot of these courses, they really kind of, they raise sort of larger conceptual issues. So I think in my teaching at Amherst, I've been prodded to make those connections, which I think has really been able to kind of serve me pretty well in kind of making some of the points that I made in this book, whereas again, I found a lot of these law professors, incredibly smart, I learned a lot from them, I'm incredibly grateful to them, but they tended to see things a little bit more narrowly, a little bit more in the weeds and maybe missing some of the larger conceptual issues at play.

- I think that is an excellent conception and summation of why I found law school disappointing after going straight from Amherst, and I expected it to just be a further exploration of LJSD classes and it was not. So let's turn to some of the questions from the audience. The first one is from Charles, who asks this. "In this day and age, in this technological black box, what is your clear definition of a fair legitimate election?

- Well, again, it's, you know, I think we have a pretty decent sense of what a fair legitimate election is. I mean, I think a fair legitimate election is one in which people who are eligible to vote are able to vote and that there isn't a politics of voter suppression, which is interfering with the act of casting a vote and that all votes that have been legitimately cast are counted and recognized. So I think we have a pretty clear standard of that. Now, obviously we've seen ways in which elections have failed to live up to that ideal. I think it is fair to say that voter suppression has emerged as something as a staple of Republican politics in the last, in, actually a pretty long history of it now, but really has kind of emerged as a staple of Republican politics. But in terms of trying to figure out what the normative ideal of a fair election, I think we have a pretty good sense of that. And maybe one other thing I could quickly add, Laura, is even though I've mentioned this is about the, about voter suppression, I should also say that American elections tend to be pretty fair, they tend to be pretty fair. And that's again one of the, there's one thing about, there's the kind of the typical authoritarian response is to say that an unfair election, totally rigged election actually is a totally fair election. That's what you have in the kind of, it's sort of like a part of a standard feature of an authoritarian regime insisting that a totally rigged election, absolutely fair. The remarkable thing about what Trump has been doing is he's basically insisting that elections that really function properly are not functioning properly, and that's a totally different kind of danger. And kind of just running down the perception of the legitimacy of our elections when they actually function pretty darn well.

- So the next question has to do with the public perception of that point. Assuming that Trump challenges the outcome of the election refuses to acknowledge the clear results, how do you see that response playing out in the public? I mean, it's hard to know that you can trust polls, but he certainly does not have the support of the majority of the country. This is a question from Mark Carol, class of '74, and he goes on to say, "Are you concerned about violence?"

- Yes, I'm concerned about violence, absolutely. I'm concerned about the kind of a messaging that issues from Trump I mean, I do think that it's fair to say that part of the key of his, or part of the hallmarks of his political style has been dividing Americans and he's an incredibly divisive figure and we've seen-

- Just given the summer of protests though, especially, I just wonder if he really digs down on this, it just seems like there could be a massive revolt from those who tried to vote him out.

- Yes, absolutely. I mean, really think, I mean, there are all sorts of things that we can observe. One of the things we can see, this kind of this overcommitment of federal agents important, we saw the way in which he was actually prepared back in that Lafayette photo opportunity in early June, we recall that he was prepared to invoke the Insurrection Act of 1807 to put federal troops on the street and as you recall, it was only, then-attorney general at the time it was Attorney General, Barr and Defense Secretary, Esper, who talked him out of doing that, and we also see the kind of dog-whistle politics that he engages in. We can even go back to the early days of the COVID pandemic when in order to score political points against democratic governors, he basically said he issued some kind of tweets saying, "Liberate Michigan, liberate Wisconsin." And we saw the way in which his supporters answered that call by protesting not simply in front of the state houses with automatic weapons, but within the state houses themselves. And we've seen, in Kenosha, we've seen violence breakout. And I do think that the opportunity to [Inaudible] violence very, are great and it's a very, very concerning thing.

- How do you not feel sort of despondent about that fact? In other words, is it too late for us to do anything about this?

- I don't think it's too late, I mean, let's put it this way. I mean, I think one of the things that I kind of said early on was even if he doesn't concede, I think we could get to a situation in which he submits to defeat that is, I don't imagine Trump barricading himself in The Oval Office on January 21st, the day after that Biden has been duly inaugurated president of the United States and surrounding him with a phalanx, surrounding himself with a phalanx of rogue secret service agents. I don't imagine that happening. So I do think if he loses decisively and again, in the ways that I've already described to you, if he loses decisively, I think he'll have no option, but to submit to defeat and I think he will be pressured to submit to defeat by members of the Republican leadership as well. I can't imagine, I mean, again, it depends on the margin of the defeat, but I do think that you will find GOP lawmakers at that point pressuring him to leave office. I can't imagine, again, I, it's not like I have any sort of idealized picture of Mitch McConnell, but I do imagine Mitch McConnell, other leading republicans basically pressuring Trump to recognize that further fight is futile. Again, that wouldn't necessarily spare us from a certain amount of domestic unrest and from some ugly images on the street, but I do think it would spare us kind of the worst type of constitutional crisis that we could see as a result of really a protracted electoral struggle come this fall.

- Yeah, and one of the things that you do in the book, I thought it was helpful was sort of explain that this is not just about Trump, right? This is very much also about Republicans in Congress and they have a huge role to play in how this goes potentially, and actually flows nicely into another question from Cole Graber-Mitchell, class of 2022, who asked, "How does Trump's norm-breaking reflect on the Republican Party's legitimacy as a political party?"

- Yeah, I think that's an excellent question because I think one of the things that we expect about norms is, as we mentioned earlier that if you break norms, you don't necessarily face a legal sanction. Norms are different from laws, but you would expect them to have to pay some kind of political price. And I think one of the most astonishing things about the last three and a half years, hasn't simply been Trump's pattern of norm-breaking, but it's the way in which he has not had to pay a political price for breaking these norms and the way in which the Republican Party has defended him and defended him aggressively. And one of the things that are worth bearing in mind is that in 2012, this is just eight years ago, Mitt Romney was the standard-bearer of the Republican Party. He was the Republican Party's nominee for president of the United States. He represented the Republican Party. Now it's hard to imagine a more marginalized figure within the Republican Party. You even have Donald Trump Jr. basically, floating the idea that Mitt Romney be thrown out of the Republican Party. And I really think it underscores the way in which the Republican Party has been deformed from a truly respectable political organization into an organization which is now willing to look the other way and not simply look the other way, but basically, aggressively defend attacks on some pretty basic norms of our constitutional democracy. It's a very, very troubling fact. And again, obviously, I mentioned Romney because he was the only one with the integrity to vote for Trump's removal during the impeachment process, during the trial in the Senate.

- You know, one of the things I often wonder about is how much that the pandemic effects that analysis, because pre-COVID, I think that you could say the GOP will line up against him no matter what, but given how disastrously this has been handled and how unfortunately so many lives have been lost, I do wonder how that affects what Republicans in Congress do in terms of how this election goes. And another student question here from Charlie, Sotheby, class of '22 has a similar theme, which is, "To what extent does the pandemic exacerbate the threats posed by the three doomsday scenarios outlined in your book?" I think the pandemic is a really interesting wrinkle to all of this.

- Yes, I agree completely, and actually going back to the point that you're making about the Republican support in light of his response to the pandemic, the one thing that I suppose we could say that's ironic about that is that when Trump engages in transparent abuses of the presidency, the Republican Party stands by him and Republican lawmakers stand by him. It's really when he engages in kind of an epic failure of leadership that we really see maybe his support kind of eroding, but they're certainly willing to look the other way when he was kind of aggressively abusing the office of the presidency. And I do also agree that the pandemic, it obviously creates a very, it just raises a host of problems that we otherwise would not have seen in this election. I think under the best of circumstances, this election would have been trying, given the character of the inhabitant of The White House, but we now know that this is going to be an unusual election and I do think that the unusual election is putting unusual pressures on our electoral system. I don't think our electoral system is well-designed to troubleshoot these kinds of pressures. And again, this kind of raises the concern that even this is kind of goes a little bit to the point that you're also making, Laura, that on one level that pandemic weakens Trump as a candidate and it certainly has weakened him. In fact, it might very well be the case that, but for the pandemic, he would have actually simply been re-elected and we wouldn't have been necessarily thinking through these scenarios. So there's no doubt that it has weakened his candidacy and made him more vulnerable at the same time and this is maybe the kind of the overdetermined quality of the danger that we're facing because he's a weakened candidate and because of the unusual ways in which voting is going to be conducted. It also plays into his kind of conspiracy theories and so it both weakens him and also provides fuel for his kind of conspiracy theories and his attacks on the integrity of the process, which ultimately makes for a pretty volatile situation.

- Yeah, I think that's a really interesting way to think about it, that duality there, it's not just all one thing. Here's one from George Fleming, who asks, "If the election is once again thrown over to the Supreme Court, do you have a serious concern that this court is too politicized now to function objectively?" I think that's a fascinating question given the dynamic between Chief Justice, John Roberts, and President Trump.

- Yes, and I would say, and just to kind of comment on the dynamic, again, it's not like I interviewed the chief justice or that I have any kind of-

- You don't have all of his clerks on speed dial?

- Right, exactly, yeah, well that I do, but I'm just saying that his personal number I only use on very rare occasions, but I think it's probably pretty fair to say that John Roberts does not think President Trump is fit to be president of the United States. Again, just a little bit speculative on my part, but I certainly think there's no love lost between those two figures.

- And you're saying that based on public statements alone.

- Yes, I would say that's based on public statements alone and also say something it's also based on John Roberts as a jurist. I think that there's no doubt that John Roberts is a very conservative supreme chief justice, it's also very clear that John Roberts is a true institutionalist. He really is concerned about institutional integrity and I don't think he has any stomach for people who trample on the integrity of institutions of constitutional democracy. And so if it were the case that for some reason, the Supreme Court had to intervene in one of these electoral disputes that could emerge in a swing state in the counting of mail-in ballots, do I think that Trump could reliably count on the Supreme Court deciding in this favor? I would say most emphatically not. I do not think that he could reliably count on the Supreme Court intervening on his behalf or deciding on his behalf. The other thing I should mention, this gets a little bit more technical, Laura, but if for some reason the electoral dispute goes on for weeks and ends up landing in the lap of Congress, which is something we haven't exactly explored but it is a possibility and it has happened in our nation's history. At that point, I would say that most constitutional experts would agree that the Supreme Court has no role to play in the resolution of an electoral dispute. So again, once it lands in Congress' lap and as it just kind of a chronological matter, that would happen around on January 6, 2021, that's when according to law, the Congress would be basically opening electoral certificates, tallying them up and declaring an official winner. At that point, the Supreme Court would have no role to play if the dispute lands in Congress' lap.

- I think we might have time to sneak in just one more question here if we could. I'm curious about what your answer will be for this one. It's from Michael Cole, class of '06, and his question is this. "If Trump loses, how can his many die-hard supporters be reassured of the integrity of our electoral process?" And I think that's such a fascinating question, because when you think about his base, you know, we described them as if there are some monolithic group. And I actually think the basis is sometimes more complicated and given that at least according to some polls and who knows whether this will turn out to be true, that the base has at least shifted from where it was in 2016 and that for a fair amount of women, in particular, have dropped off in ways after voting for Trump in 2016. I am curious what your thoughts are on what his base of support does with a contested result.

- Yeah, I think it's a really important question and I think obviously one of the things that has both facilitated and empowered the rise of Trump is this incredibly fractured media environment in which we operate, so whereas we all wish that our only source of information was you, Laura, and your colleagues at CNN. We know again that we have these alternative sources of news.

- I'll tell Walter you say that.

- Right, and unfortunately, if your only source of information is these alternatives, these new sources, then you very often hear a very, very kind of different message. So I think one thing that would be absolutely critical is for leading Republican lawmakers and again, this might be again, aspirational, but for the Mitch McConnells of the world to kind of step forward and say, this election was conducted properly, I see this result as legitimate, the president needs to step aside. And again, I think the chances that you'll find in Mitch McConnell and his ilk messaging in that way are greatly increased if Trump experiences a decisive loss. And it's only really if things turn, as I said, on the late count of mail-in ballots and swing states with now margins that we really could see, a much more, much more troubling scenario spell itself out. But I do think we need to hear not just "CNN" voices or "New York Times" voice, we need to hear Republican voices and very powerfully coming out and saying, "This election ran in the way elections are supposed to."

- Well, I think there was even one op-ed in "The Washington Post" yesterday from a very prominent GOP lawyer, sort of laying out the case that all of this talk of a rigged election and widespread voter fraud in mail-in ballots is basically completely unfounded. So it'll be interesting to see as we get closer and closer to the election day, whether people feel a greater sense of urgency to speak out on the heels of all of these tell-all books and more and the president ramping up some of his own rhetoric getting closer to the election. Thank you so much for taking the time to do this, Professor Douglas, it's been such a treat to talk to you about this book. Everyone should read it, it's terrific. It's so helpful I think to lay out all of the different scenarios. I found it terrifying in many ways to think about how long this could go on, but it's super instructive and very easily digestible. So I encourage everyone to read it and it's been such a pleasure talking to you and seeing you today and so I will hand it over now to Dean Epstein.

- And again, and thanks again for a fantastic conversation, Laura, and again, it's wonderful to see you.

- Thanks.

- And thanks again, Katherine, for the very lovely introduction.

- Well, so thanks to both of you, this was absolutely wonderful and fabulous. Just the words that Lawrence used, Laura, your questions are so penetrating and analysis, so intriguing, and Lawrence, you explicate everything so well. So it's really kind of liberal arts learning and teaching in action and I'm really thankful for that. Before we close, I'd like to note the next Amherst Conversation. It features anthropologist and obstetrician-gynecologists, Carolyn Sufrin of the class of 1997 and her Amherst mentor, Deborah Gewertz, professor of anthropology. The two will discuss anthropology, medicine, and reproduction in the prison system and the conversation will focus on the intersection of mass incarceration, the politics of reproduction, and systemic racism. Please join us for that event on October 5th at 4:30 p.m., and now thank you everyone, and have a great day.

Betsy Cannon Smith (00:00:09):
Good evening and welcome. Thank you for joining us tonight. I'm Betsy Cannon Smith class of 1984, chief advancement officer at the college tonight. I have the distinct pleasure and I've actually had the pleasure over the last 30 minutes or so to be chatting with Austin Sarat, the William Nelson Cromwell Professor of Jurisprudence and Political Science at Amherst College and associate dean of faculty. 46 years ago, Professor Sarat joined the faculty at Amherst College and in the ensuing five decades, close to 8,000 students -- yes, you heard that right, some of you are out there -- 8,000 students partook of at least one course from Professor Sarat. Some of you on the phone with us took multiple courses and some are looking forward to your first course with him this year. Now the good news for those of you who've experienced a Sarat class is he can't call on you tonight.

Betsy Cannon Smith (00:01:01):
This setup doesn't allow it, but believe me, he wants to, he's been looking at the list and saying who he'd call on if he could, among the memorable course titles, and some of you will remember these: Murder; Secrets and Lies; America's Death Penalty; the Social Organization of Law; The Meaning of Catastrophe, and this summer Disaster, Catastrophe and Democracy, something he'll tell us a little bit more about in the hour to come. Professor Sarat holds six degrees, two from Providence College, his undergraduate and an honorary degree, an M.A. and Ph.D. from the University of Wisconsin, an honorary degree from Amherst College, and a J.D. From Yale University Law School, which he undertook while teaching full time at Amherst. The record of his publications, if I read it will take the entire hour, is an extraordinary year by year account of a curious mind at work. To have taken a Sarat class is to have experienced the intensity of the academics with the passion of a teacher who cares as much about the experience of the classroom, the partnership in that classroom, and the exchange of ideas as he does about the topic.

Betsy Cannon Smith (00:02:19):
We'll hear more tonight about how he translates that in-person classroom experience to a virtual experience. Something he's already had a little experience with and is preparing to do for the fall outside of the classroom. Professor Sarat does the important work of innovating when he sees a need on the campus, of collaborating with colleagues. He is one of the college's top cheerleaders, supporters, and just a person who will always stand beside his colleagues and his students. He'll throw an occasional song and dance in if you're lucky seeing him walking across the campus or down the hall, and in fact, was singing just before we came on the air tonight. Tonight, he'll share what it's been like to teach through the pandemic, to work in ways we've never worked at Amherst College and to prepare for the fall where approximately half of the students will be here in Amherst on the campus and the other half will be scattered across the globe. Tonight, Austin and I are on opposite sides of the campus in our respective homes, both in Amherst, but doing what we'll do through the coming year through these screens. So we hope you enjoy the conversation tonight. I encourage you to ask questions. There's a button on the screen where you can submit those and I'll remind you that this is being recorded. Let me turn it now to my friend and colleague Austin Sarat, and ask Austin, Austin, are you looking forward to the fall?

Austin Sarat (00:03:44):
So first of all, thank you very much for the introduction. That was, that was incredibly, incredibly kind, good judgments. I guess I should stop now because that introduction was such a real setup. Betsy, are you looking forward to the fall?

Betsy Cannon Smith (00:04:04):
Am I looking forward to the fall? I am because we have students coming back and I am really looking forward to seeing how our faculty present the students with the courses that they've selected and work together with students who are in places all over the world.

Austin Sarat (00:04:21):
So, I'm really quite excited about the fall and really ready to go. I feel like a kind of bullpen pitcher waiting to be called into the game. I'm really excited about the semester because over the last 10 days or so, I've been connecting with students in various ways. My advisees, students from the summer bridge program, and a really special group of students who are going to be taking my first-year seminar and the opportunity to talk with Amherst students, to hear about what their interests are and to hear about what they're looking forward to. I find incredibly bracing and incredibly exciting today. I talked to one of my advisees, her name is Dania, and I asked her, as I often ask my students, not what they want to do after college, but who they want to be while they're here.

Austin Sarat (00:05:18):
And what were her dreams about her college education? I said, I want to hear about your, I want to hear about your dreams and what Dania has said, she said, you know, Professor Sarat, I'm really interested in the sciences and I really want to take a lot of STEM courses. And I said, Dania, that doesn't sound like a dream. That sounds like a plan. And she said, well, the dream that's associated with that plan is that I will end up using my skills and science to improve the community in which I live. Now, over the weekend, I talked to another student, who's going to be in my first-year seminar. And I asked the very same question. I said, what is it that you're thinking about with your college education? What is your dream? And he said a very familiar thing. He said he's looking forward to making lifelong friends.

Austin Sarat (00:06:12):
And I then asked him, well, is there anything else? Is there anything related to your learning experience in the classroom? And he said, well, he said, the thing that I'm really looking forward to is learning to be a better listener. And a third student that I talked to over the weekend, and again, what I'm doing is I'm connecting via zoom, with each student in my first-year seminar, individually, to get to know them, to introduce myself. I have to say they're enormously poised. They seem not at all nervous about talking to me. And the truth of the matter is I would be a little nervous about talking to me, but they don't seem to be at all. But this student, again, when I asked, like, what is your dream for your college education? This student said that they were most looking forward to being on campus in difficult times.

Austin Sarat (00:07:14):
And I was just blown away by that. So I'm blown away by a student who thinks about her STEM education in terms of what it is that she can do for her community. I'm blown away by a student whose dream about his college education is learning to be a better listener. I mean, that is completely, that is completely fabulous. And I'm blown away by a student who's excited about coming to a campus in difficult times. You see, I've always believed about Amherst and I believe it more now than I have ever believed it, that Amherst gets it right in the way in which it treats students. It respects them and it trusts them. And this upcoming semester, more than any in my time at Amherst, we're going to see the way in which our students reward that trust and the way in which they earn that respect. So I can't wait for the semester to start. I'm ready to go. I've sent the students their first assignment and have been thrilled, have been thrilled to connect with them.

Betsy Cannon Smith (00:08:34):
So, Austin, I want to hear about when we transitioned in March, but I want to ask you first, you asked the students and I think this is something you asked most of the students you work with about what their dreams are and what their hopes are, so I'd like to know what your dreams were when you came to Amherst in 1974 and what your dreams were as a professor and how those have played out. And what, how this time has affirmed those or changed those.

Austin Sarat (00:09:04):
So, my dream, Betsy, was to come to a place that would allow me to grow intellectually. That would want me to grow intellectually. My dream was to come to a place where I would be challenged by my students to be the best teacher and the best thinker I could be. My dream was to come to a place in which the disciplines that we taught didn't matter as much as the subjects that we were trying to elucidate. And I found that dream at Amherst. It's not always been an easy place for me to be. Amherst standards are very high, and they're very high across the board and they're high for our students. And they're high for a wonderful staff and they're high for our faculty. But I found that dream here, and the dream that you've asked me to elucidate is as powerful today as it was when I arrived. 46 years down and I want 46 more to go. The Amherst classroom is paradise. In my paradise students don't get A's, but it's still a paradise because I think I am doing what Amherst wants to do for all of its students. I'm holding my students to high standards and I'm surrounding them with care because that's what Amherst promises. We will hold our students to very high standards and we will care for them and care about them as human beings. And it's a pleasure 46 years in to still be able to do that work.

Betsy Cannon Smith (00:10:53):
Thank you. So let's talk a little bit about March. Let's talk about that day when the decision was made and you went from thinking, you could walk from your house to your classrooms, see your students, do a little dance, do a little entertaining and be face to face. And then overnight we went remote. Tell us a little bit about that.

Austin Sarat (00:11:15):
So it was obviously nothing that anyone had prepared for and Amherst made the right decision then to send students away from the campus to protect the health, their health and safety, and the health and safety of the community. And Amherst is making the right decision in the fall to bring back students onto the campus. And we have a really great, a great plan to do that, but the spring was really a test for me and I think all members of the faculty. We say what we want to do with our students is we want to teach them to be resilient, flexible, creative, and persistent. And the spring was a test about whether I would be resilient, flexible, creative, and persistent, and frankly, I didn't wish for the challenge, but I'm glad I had the challenge. I think I'm a much better teacher today by virtue of what I learned last spring.

Austin Sarat (00:12:22):
I think I'm a much better teacher today by virtue of what I had to do to remain connected to my students last spring. Now, again, let's just remember that last spring I was teaching two relatively small classes. I was teaching a class called Murder as an intensive writing class, a class in which students write for every class. They get responses. We workshop their writing and they meet with me individually. And I was teaching what we call a research tutorial called America's Death Penalty in which a small group of students was working with me to become collaborators with me on a research project. So those were the classes I was teaching and zoom worked for me. It's very possible in zoom to point to someone and say, Betsy, I have a question for you. And then to say Charlie, what do you think about Betsy's response?

Austin Sarat (00:13:19):
But you had to do certain things to prepare. And I did certain things to prepare. I connected again with every student individually over the spring break to check in with them to find out how they were doing. I sent them a zoom protocol. I sent them, here are some rules for being on zoom. For example, I said, it'd probably be better not to appear on zoom in your pajamas, or if you have a nasty sign on the wall about how much you dislike Professor Sarat, it would probably be a good time to take that sign down. I reminded them that one of the real problems of zoom is that people stare blankly into the camera. They don't acknowledge each other. They don't nod and smile and say, you know, hallelujah and amen to what people say in a classroom.

Betsy Cannon Smith (00:14:08):
Amen.

Austin Sarat (00:14:08):
Way to go, Betsy. What a zoom classroom lacks is it lacks the so to speak energy of the crowd, it lacks the chemistry of the connection.

Austin Sarat (00:14:19):
So had to work to try to make that chemistry work. I would break up the students, even in these small seminars, into so-called breakout groups and send them off to talk with each other about a problem that I gave them. And I assign them to groups and gave them group projects which they would do over every weekend. And part of the group project had a social component to it. I would ask them to choose a film or to choose a music video that best exemplified what we were going to read and talk about in the next week. But it wasn't all roses. So it's very clear that zoom learning is very different because of the lack of the energy of the crowd. Zoom teaching takes more energy from me. And it was just, that's harder, an hour and 20 minutes on zoom is a long time. Students face challenges.

Austin Sarat (00:15:21):
What do they face? They face the challenge of distraction. So you have to say to students, don't look at your cell phones, don't check your email. They face the challenge of loneliness, of being alone. They face the challenge of dare I say of infantilization. They go from being on a college campus where they were independent to being back for many of them, not for all, but for many of them being back in their families' homes, and frankly, we were all scared. The world felt like it was just coming undone. What I think my students appreciated was they appreciated a visible commitment to them and their success. When it had to be a little bit more flexible and understanding in terms of the learning challenges and when I had to help students. So I hope some of our listeners who are interested in what it's like to be a student in a zoom class will ask for some advice, cause I got lots of advice for them, but again, I provided some scaffolding.

Austin Sarat (00:16:37):
I urged them to do studying over zoom, to arrange at eight o'clock at night, with four or five of their friends, just to turn on the zoom and to study together just the way they would have done in frost library. So that was kind of the spring experience. I have to say that I felt that my students needed me more than ever. And I have to confess I needed them more than ever. One of the students said at the end of the semester, in our last class, how much she had appreciated what she called the touch of the normal world that our class represented. And that sense we're holding onto something that gave us a touch of what was, for us, for the students and for me, the things that sustain us in the normal world. That was a big part of the spring experience.

Betsy Cannon Smith (00:17:43):
Austin, you said that you have a lot of advice for students who want to learn how to be successful or engage

Betsy Cannon Smith (00:17:50):
this way. What advice did the students offer to you about your teaching that may have either informed this spring or will influence how you go forward in the fall?

Austin Sarat (00:18:02):
Sure. The advice they gave me was to keep on doing what I was doing. One of the students rather irreverently referred to my method as what she called the Saratic Method. And she said the Saratic Method keeps students on their toes in zoom, just the way it does in the classroom. She said the only difference, she said, in her experience of the classroom was that, and I love this phrase, she said we didn't get to breathe the same air that you breathe. Now, I didn't know whether that was a relief because they weren't breathing the same air. The other thing that students wanted, I believe, is they wanted faculty to be visible in the commitment to their success. They wanted faculty to reach out to them. They wanted faculty to connect with them. They wanted faculty to realize there's a difference between being in a classroom, on a college campus, where everybody is in it together and being home alone, studying and confronting those moments of doubt. Like, I can't understand this reading. It must be, it must be me. It can't be anything else. So the effort to connect with students to stay in touch with them, send them emails, just checking in. How are you doing? Those are things that I think we all do when we're teaching on campus, but we all really need to do when we're teaching, when we're teaching virtually,

Betsy Cannon Smith (00:19:39):
Can you talk a little bit about the summer both about your own preparations and I think you snuck into those preparations by taking on teaching a special summer course, but then also talk a little bit about what the faculty as a whole, what the faculty have engaged in, in terms of getting ready for the fall.

Austin Sarat (00:20:00):
So I'm proud of a lot about Amherst. I don't think I've ever been prouder of the institution in terms of what I've seen over the last six months. And that goes for the staff, that goes for the people who continued to go on campus to make sure that the campus would be running for the 150 or so students who were there, that went for people in the admissions office that had to kind of reinvent an admissions process, that went for people in advancement who made sure to keep connected to our constituencies, our families, the parents of Amherst and our alums. But I also have to say I'm really especially proud of the faculty. More than 90% of the faculty have spent the summer retooling their courses for the online world, meeting weekly, to hear about tricks and best practices in online teaching, and to talk with each other, to become communities of, communities of learners.

Austin Sarat (00:21:11):
We talk with each other about what we were going to do, what we thought would work, and what we thought would not work. And that work will conclude in the next, that that kind of intensive preparation over the summer will conclude in the next couple of weeks or just in time for the beginning of the semester. The faculty of Amherst, I think more than ever understand that the primary role of the teacher in an Amherst classroom is not to be the judge of the students, but it is to facilitate their success, to push them beyond capacities that they don't know that they have. You see the life of an Amherst classroom is the life of becoming, not of being. We imagine and see what our students can become. And we ask them to reach for that thing. And what faculty have done over the summer is to learn some new techniques to keep our classroom this place of becoming. Some of our colleagues

Austin Sarat (00:22:28):
and I'm going to mention some names, have, I guess, to use the title of a song from Hamilton, just blown me away. Chris Durr in chemistry and Marc Edwards in biology and the wonderful work they did this summer in what they call the summer incubator project for rising sophomores who want to work in STEM. My good colleague, David Hansen who's pedagogy has I think, been reinvented to meet every student where they are. My colleague, Adam Sitze, whose dedication to getting it right, making sure that students understand that getting it right is the first commitment of an intellectual, these are faculty who spent the summer taking those commitments and learning how to make them work in a virtual world. And if I may, could I tell you a little bit about my disaster course?

Betsy Cannon Smith (00:23:35):
Absolutely. Please do. And before you do, I think this relates to a question that J.J. Gertler has just sent in, which is, has the pandemic given you issues that can be used in classes or inspired topics for new classes? And I'm thinking maybe that your summer class had something to do with that.

Austin Sarat (00:23:52):
I love J.J. J.J., that was a great question. That was in fact, my summer class. So a few years ago with my good colleague, Lawrence Douglas, we taught a class called the meaning of catastrophe. Having taught it for several years. And again, I have to acknowledge something. I hate the end of the semester. I miss the life of the classroom. Some of my colleagues, they find relief when the classes are over and they can turn to their research. What I find is I find loneliness. I miss my students and this year I knew I was gonna miss them more than ever in our isolation and in our fear and in our uncertainty. So I reached out to students that I've had in a variety of classes. And I said I have this idea. Here's my idea. I want to teach a class with the happy title, Disaster, Catastrophe, and Democracy.

Austin Sarat (00:24:53):
And it's not for credit. You're not going to get any credit for it. I'm not going to get any credit for it, but we'll meet twice a week, just like a regular Amherst class. I'm going to send you a syllabus that is going to be every bit an Amherst syllabus filled with complicated readings, and we're going to meet twice a week and we're going to have a zoom class on Disaster, Catastrophe, and Democracy. My class began with the following question, Betsy, I wonder whether you might help me answer this question. I started my zoom class this summer, Disaster, Catastrophe, and Democracy as I do most of my classes. I come into the room or I entered the zoom and I say, I have a problem. And then I turned to a student, in this case, Betsy, and I said, Betsy, here's my problem. I want you to imagine, God forbid, heaven forbid that the bridge connecting Hadley to Northampton over the Connecticut River, that that bridge collapses. It collapses all of a sudden, there's no one on the bridge. And for two weeks the army corps of engineers and someone else builds a kind of pontoon bridge over the Connecticut. So people can get to Northampton. Why they would want to go to Northampton I don't know, but here's my question. Would you regard that bridge collapse as a catastrophe?

Betsy Cannon Smith (00:26:26):
Oh, disaster, catastrophe. Would I regard it as a catastrophe? I think I regard it as an inconvenience, but not a catastrophe if there were not if the bad outcomes were delays and things being put up, you're gonna come right back at me with something that tells me it really is much more of a catastrophe that I realize. So this is my payback for not taking a class with Austin Sarat, I'll let you all know.

Austin Sarat (00:26:54):
So, Betsy, I want you to answer the questions that I ask you. So here's the next question. If there was someone on the bridge when the bridge collapsed and that person was killed, would you regard it as more than an inconvenience?

Betsy Cannon Smith (00:27:11):
Yes, absolutely.

Austin Sarat (00:27:14):
Why? Why would you regard it as something more than an inconvenience?

Betsy Cannon Smith (00:27:19):
Because a person's lost their life due to this.

Austin Sarat (00:27:24):
Good. And would that loss of life qualify for you, would it make the event a catastrophe?

Betsy Cannon Smith (00:27:31):
Yes.

Austin Sarat (00:27:33):
The loss of a single life.

Betsy Cannon Smith (00:27:37):
I think the loss of a single life is catastrophic to the people who have, for whom that person is part of their lives.

Austin Sarat (00:27:45):
The question I asked, that's not the question I asked you. I'd like you to answer the question I asked. I didn't ask you, I didn't ask you whether or not it would be catastrophic to that person.

Betsy Cannon Smith (00:27:58):
Okay. So that would probably be a disaster because catastrophe seems to have a much larger impact.

Austin Sarat (00:28:08):
Good, fabulous. Suppose the person who was killed was the only person that knew how to make a vaccine for COVID and they're killed and with them goes the vaccine and the possibility of having a vaccine for COVID. Now, the purpose of my asking you these questions is just to really, in a sense, exemplify what can be done over zoom. So that kind of back and forth in which I asked students to begin to develop a conception of what the catastrophic involves. And I insist that they listen and answer the question that I ask them. Are you remembering what I started with the student whose dream is to come to Amherst and become a better listener? Well, part of our job in the Amherst classroom is to show that we listened to students. If a student says, Oh, X group is oppressed, we might say, well, that's interesting, but are you sure it's oppression as opposed to disadvantage?

Austin Sarat (00:29:16):
And we might then ask them, well, let's play out the distinction between disadvantage and oppression. So my class on Disaster, Catastrophe, and Democracy, went away. My class in the spring. When we did zoom classrooms, we met twice a week. I would zoom with them individually. I assigned them to zoom, to zoom groups. And we did what again what we would do in any Amherst classroom. An Amherst classroom is a chance to take students -- are you ready for this --on a journey from the familiar to the strange. So we start with an example, like a bridge collapse, and students think, well, they really understand that they get that. And then what we want to do is we want to complicate their understanding. And we say to them, look, you're really not going to be able to answer this question until you've read Hobbes and Rousseau and Voltaire on the Lisbon earthquake and Primo Levi on the Holocaust.

Austin Sarat (00:30:17):
And when you read legal cases, like the Korematsu case about Japanese internment or the steel seizure case about presidential emergency powers. And that's what we did in that class. And it started on the 22nd of May and we extended it for a week. The students didn't want it to end. We extended it for a week. And I must say in the course of this class, I did something that I almost never do. So I'm the kind of teacher that mostly doesn't like visitors in my classroom. I don't mean, I don't mean alums or new students. They can come, but I don't bring speakers into my classroom because the work of my classroom is the work of dialogue and listening. But I thought I want to reward these students. I've got nothing to give them. So, and this is another thing that is remarkable about Amherst,

Austin Sarat (00:31:15):
I decided what I wanted to do was I wanted to invite some prominent public figures to come to my class and to participate in a question and answer with my students. And we had Kevin McAleenan in, a fabulous Amherst graduate who was acting secretary of Homeland security under President Trump. We had Juliette Kayyem, who was in the Department of Homeland security under President Obama. We had Senator Chris Coons. We had Congressman Jamie Raskin. We had former mayor Rahm Emanuel, and we had mayor Pete Buttigieg. Now my students thought, Oh my God, you know all these famous people. And of course, I didn't want to tell him, and I shouldn't say this, but of course, I didn't know those people. I reached out to Amherst alums and asked alums whether they would be, they would spend some time making it possible for this group of 16 Amherst students, to whom those alums had no particular connection, to meet with Rahm Emanuel and Jamie Raskin and mayor Pete and the alums did what they always do, which is, they stepped up.

Austin Sarat (00:32:39):
So I found the class to be remarkable. Remember what I started with Amherst does it, right? It trusts and respects its students. And my respect for students, 16 students for no credit, there were no exams, they came to class well-prepared. They were well prepared for my questions. They made me work really hard to get it right, to use that phrase. And I loved, I absolutely loved working with them. After finishing that class I taught a comparable class for high school teachers. I assembled 16 high school teachers from around the country, and we did a month on Disaster, Catastrophe, and Democracy, and my July teaching ended with two weeks of working with some of the most remarkably interesting, most talented students. And those are students that participate in our summer bridge program. I taught a class called Justice and Power in our summer bridge program.

Austin Sarat (00:33:52):
And in each of those occasions, I did what I described earlier. You guys got to connect. We got to, you gotta study in zoom, in zoom sessions, we gotta do breakout. You gotta do breakout rooms and I would meet with them. I would meet with them individually. So I'm, as Barack Obama would say, fired up and ready to go in large part because of the work that I did with students over the summer and having taken the training course, this wonderful training course, that was made available to faculty.

Betsy Cannon Smith (00:34:34):
Have you been able to continue your research or was really the development of that course more a part of your research?

Austin Sarat (00:34:41):
So you, you asked me about my dream about Amherst, and part of my dream when I came was that I would be in a place in which teaching and research would not be seen as in opposition and Amherst has been that place for me. Yes. I've continued my research. I'm doing a project with four Amherst undergraduates called the fate of lethal injection, which is looking at what's happened to lethal injection as a technique for execution in the United States. Over the last decade, I've worked with a student, oh this extraordinary student whose name is Ryan on a project called The Death Penalty in Crisis. And we're looking at what's happened to the death penalty in times of crisis in the United States, in pandemics and recessions, and in a World War. I sent off a book to the press called Law's Infamy, and I've done something that I felt called to do

Austin Sarat (00:35:58):
and that is I've spent a fair amount of time writing for a public audience, writing opinion pieces published in places like the Guardian or Slate or a site called The Conversation or a site called The Bulwark which I really quite love because it has a broad audience writing about the fate of American democracy, the crisis of American democracy. I published a piece today called Memorializing Miscarriages of Justice, in which I reflected a little bit about what it means to write in dangerous times. What is it that the vocation of being a writer speaking in dangerous times, what is that about? So I found the teaching, as I have always found it, to be energizing for my scholarship, and I have loved and continue to love working with Amherst students on significant research projects.

Betsy Cannon Smith (00:37:06):
You talk about these opinion pieces, Austin, was there a moment or an event, perhaps a disaster or a catastrophe, that inspired you to take that turn and really start doing a lot more of that public writing, writing for the broader public?

Austin Sarat (00:37:24):
So I think that intellectuals and academics have an obligation to try to repair the world. I think that we discharge that obligation largely in our classrooms by equipping students to think well, by equipping students to listen carefully, by prodding our students to imagine a better world, but I felt if you will, the call of repairing the world and it's what I do, I write. So yeah, that's kind of what I have done. What really got me started on this public writing was reading a variety of studies about how it is that Americans have begun to disinvest in democracy. So in response to a national survey, people who were born in the 1930s, that it means people who are now well into their eighties when they were asked, do you believe it is essential to live in a democracy, about 75% of them,

Austin Sarat (00:38:38):
this survey was done in 2015, about 75% of those people who were born in the 1930s, said that it was essential to live in a democracy. People born in the 1980s, that is, say people now coming up on their 40th birthday, same survey, just broken out by birth cohort, 25%, 25% said it was essential to live in a democracy. And reading that study, I thought I've got to say something. And I wrote a piece and it was published. It was published in The Guardian and it got a good response. And frankly, that fueled my sense of, I wanted to use what I know to help inform public debate. And again, I have to confess, this is a situation of some tension for me because I've worked very hard, very hard so students don't know what I think in the classroom. I don't tell them this is what I think about this subject, or this is what I think about that subject.

Austin Sarat (00:39:46):
For example, Betsy, I wouldn't have answered the question for you about whether the bridge collapse was a catastrophe or not. I might've asked you to think about whether or not there, a better description might be a tragedy, even with the death. And I might ask you to think about the distinction between a tragedy and a catastrophe, but I wouldn't say what I thought. But in my public writing, of course, I do say what I think, and I have to say writing for a public audience, I'm sure is a kind of surprise to many of my teachers, my own teachers. When I was in high school, I remember taking a civics class and I didn't do very well in that class. And my teacher, I remember saying to me that she didn't think I had much of an aptitude for the subject of civics or politics. 46 years later, I guess revenge is best served cold.

Betsy Cannon Smith (00:40:47):
Well, let's shift to the faculty now. There's a lot of talk of how this remote world and what it's like for the students to be on the other side of their computers and interacting with all of their professors in this way. But we have a great question about how has the virtualization affected relationships on the faculty and is there a threat to that really critical cross-pollination that happens on the campus when faculty from different disciplines work together?

Austin Sarat (00:41:16):
That's terrific, that is [inaudible]thread is not the word I would use. So again, I want to come back and just let's talk reality. The last, you know, the last six months had been months of fear, months of disorientation, months of loneliness and months of internal struggle for lots of people on our faculty, on our staff and in our community, things that we used to take for granted that were easy now are less easy, but what faculty have done to keep connection is they've created virtual reading groups. So there was a reading group presided over again by one of our wonderful and really most inspiring teachers, Jyl Gentzler on Middlemarch, and faculty got together and they had a zoom reading group. I created a reading group with the alluring title Lawish. So we have a lawish reading group for people who are interested in lawish subjects.

Austin Sarat (00:42:26):
We have a reading group that I participated in on free speech on campus. So I think the connection cross-discipline, I think the ability to use that phrase to cross-pollinate will continue. Faculty again in their loneliness, they crave companionship and the zoom has provided it. And I have to say, again this is a story about one of the students I talked to, I was apologizing to this student who's going to be in my first-year seminar about this is going to be an online class. And I said, you know, I'm 72 years old. And I'm in a so-called at-risk group. And as much as I'd like to be on campus, I'm not going to be on campus. And this student, who's nothing short of a gem

Austin Sarat (00:43:26):
and a mensch, said to me, don't worry, Professor Sarat, I'm getting goosebumps when I retell the story, don't worry, Professor Sarat, we, meaning the students, will find a way to welcome you back to the campus. And I thought, Oh my God, Oh my God, they are going to find a way to welcome me back to the campus. So connections will continue. Some of them will be virtual and my students will welcome me back to the campus. I did say to him, maybe we could meet on the football field and you could stand in one end zone and I'll stand in the next.

Betsy Cannon Smith (00:44:07):
So I'm being overwhelmed by questions from people who are saying, can alumni sit in, can parents sit in? Can people join these virtual classes that are happening, yours and other faculty members, the actual classes.

Austin Sarat (00:44:24):
The classroom space is sacred. (Absolutely.) What makes a class work, and it's going to be harder to create it, but if we're going to succeed in creating it in the virtual world, what makes a class work is the chemistry of the classroom, the students' love for the subject and their desire to be together talking about the subject. Learning, this is terrible, this sounds like something from the inside of a Snapple cap, learning, I say, is a team sport. So I love alums, not equally. There's some I love more than others. I love parents. But I think, no, I think not in the actual classes. I hope that the advancement office will ask me to teach a class for parents and alumni. By the way, I have a class I'd love to teach. It's called Disaster, Catastrophe, and Democracy.

Austin Sarat (00:45:32):
I'd love to teach that class for alums and parents so long as they do the work. And I think you see, we will be better teachers, better teachers for what we do virtually. We will be better teachers because we'll learn new things. You know the best thing I did this summer was I took a class about teaching online, online. So I experienced what it's like to be a student sitting in my, you know, living room alone, not having anybody nod or smile when I say something, which I thought was particularly insightful, but they just kept staring straight ahead. And I felt the, I experienced that disconnection of that moment. So I think we're all gonna end up being better teachers, more inventive, more creative, more resilient, more responsive to our students. And we're going to bring some new tricks back into the classroom when we're all back on campus together. And I'd like to try out some of those tricks with alums and parents in the online world. So I guess the only thing I could say is bring it on.

Betsy Cannon Smith (00:46:53):
Oh, I think those are some marching orders there. And from the response we're getting, I absolutely think the alumni and parents will benefit from the work that the faculty have done this summer because it does give us new opportunities.

Austin Sarat (00:47:07):
It absolutely looks, it absolutely does. And let's tell the truth. The students are so excited to come back to the campus. They are so excited to come back to the campus. They're proud of what they're being asked to do. They understand they're being asked and there's some people who think they can't do it. They're 16, they're 18, they're 19, whatever it is, they cannot do it. Well, I have no doubt that Amherst students are gonna make mistakes, but remember Amherst gets it right with its students. It respects them and it trusts them. So it's going to be thrilling to have them. It's going to be thrilling to have them back. It's going to be thrilling for me to be able to teach zoom. And I hope to have office hours on campus and meet them at the football field. And I would love it if some alums have wanted to do an online class with me especially those alums, that got, let's say A grades in my classes

Betsy Cannon Smith (00:48:19):
I think one of the wonderful things about our back to school plan for this year is that over the course of the year, all the students will have the chance, the opportunity to be here either first or second semester. A few questions that I've received have been along the lines of how will you and other faculty members work to unite the remote and the in-person students who might, who will be, not might be, who will certainly be sharing a classroom, a space, an online space. How will those students who right now are remote and who'll be here next semester and vice versa, how will that experience feel? What will we do to make them feel equally part of this grand adventure?

Austin Sarat (00:49:04):
Well, that's, you know, that's going to be, that's a work in progress, but again, I'm going to just going to [inaudible]for one first-year seminar faculty member, I've already started that work. So I broke my first-year seminar up into three groups. They're called how's this for fancy titles, group A, group B, and group C. And I assigned one student to be the captain of group A and the captain of group B and the captain of group C. Those students, those groups are mixing up. Some of the students, most of my students are going to be on campus, but a few are not, those groups are mixing up the on-campus and the virtual learners. They're meeting this week to begin to get to know each other, to talk about their hometowns and where they are, where they came from.

Austin Sarat (00:49:48):
I've asked them to do what again I asked my students in the spring to do, to come to those meetings with examples of things in popular culture that best represent a secrecy and deception. So they will be equally integrated into all of the work that the seminar does. And I wanted to say something else. So I've done for the last several years what I call a first-year seminar team project. And that is, I've worked with the chief student affairs officer, the director of the Loeb Center for Career Exploration and Planning, and the Director of the Counseling Center in a team. And we meet with my first-year seminar students every other week in the evening. So it's not for the work of the seminar. And what we do in those team meetings is we talk about what it's like to be a student, what it's like to be a college student.

Austin Sarat (00:50:52):
We asked students to think about their sense of belonging. Do they feel they belong at Amherst? What are they worried about in terms of do they belong here? And by the way, that's a question that one wants to ask alums to remember what it was like when they first came to the campus, did they feel that they belonged? We asked students to talk about how they're going to balance their curricular extracurricular and social life. This year we're going to ask students to talk about what it's like to be learning remotely, because even students who are on campus, a large number of their classes, or some number of their classes are going to be remote as well. And the student affairs office also is very aware of the need to program for students who are learning virtually.

Austin Sarat (00:51:46):
So I think the campus is geared up to make sure that the 700 or 800 students Amherst students who are learning virtually remain well integrated into the campus and we're going to do it in the social programming that we do. I know that the athletic coaches are reaching out to athletes who are not going to return to campus to make sure they are part of the team. And I know faculty members are also going to be reaching out to those students to make sure that they feel that they are authentically Amherst students.

Betsy Cannon Smith (00:52:26):
Austin, how about a couple of comments about the differences between the smaller seminar-type classes you're teaching and the larger classes and how you think that these same comfortable and engaging conversations can happen in those classes?

Austin Sarat (00:52:41):
So it matters a lot if you can see everybody on one zoom screen, if you have to go to a second zoom screen, it's a little bit of a different experience. What we've learned over the summer, of course, is there ways of dealing with that. So we've learned for example, that there's this thing called zoom fatigue so that what we want to do is we want to use zoom time wisely. That means larger classes are going to do a lot of things asynchronously. We've learned for example, again, you're going to say Austin, this you had to learn? That less is more? That shorter videos work better than longer videos? We've learned the virtue of discussion groups. So even larger classes. I'll give you an example, again, one of our really spectacular colleagues is a colleague named Jonathan Obert.

Austin Sarat (00:53:30):
He teaches a class called American Government, that class regularly enrolls 50, 60, 70 students. So what we're doing this year is that class will meet in a variety of formats. It'll meet in a lecture format in which Professor Obert meet with students synchronously, but also use asynchronous material to deliver lectures. And then there will be seven separate sections of that class taught by professor Obert and other members of the political science department. And those will be discussion-oriented, those will be discussion-oriented classes. And some of the, again, the most inventive, inspiring examples of online teaching have come from colleagues in the STEM fields with larger classes. They've found ways to do virtual labs. They found ways to get students as close as possible to the in-person experience of what it's like to learn a lab science. So we've been well equipped. We've worked with a company called 2U, which is one of the best providers of online education. Everybody who's teaching a class, I believe, with 35 students or above has worked with 2U all to learn techniques for making a larger classes work in the way that Amherst classes always work. And that is to produce a kind of intense, engaged, and hopefully inspiring learning experience.

Betsy Cannon Smith (00:55:05):
All right. And how about this. Can you imagine when we're back on campus whenever that is, and you're in the classroom, what kinds of things will you, do you imagine taking with you from this that you could actually carry over into changes and things we're doing? I mean, the January course, I think, for example, the fact that we're offering a course in January. Say a little bit about both that, the January course offering, and how you might be influenced by something you're doing now and think, huh, that'll be part of what I do even back in the classroom.

Austin Sarat (00:55:38):
So I've heard, I'll start with the latter. I've already mentioned some things that I will do when I'm back in the classroom that I didn't typically do. So I didn't typically assign students to study groups like you should, you should actually study together. I expected that they either would, or they would not. My plan is to do that. And I didn't get students. I didn't gather students, get students to get together outside of the classroom before the classroom even started to get to know each other. I think I will do that. I've fallen in love. Isn't this really terrible, it's a terrible thing to confess that I have fallen in love with zoom, but I kind of have fallen in love with zoom (You do look kind of happy.)

Austin Sarat (00:56:22):
One of the features I love in zoom is breakout rooms. So you can push a button and the student, it's like an episode of StarTrek, right. I beam the students up into an alternative reality, these breakout rooms, and then I can drop in on the breakout rooms and listen to the students grapple with the kinds of problems that I ask them to grapple with. How can an idea of political democracy coexist with the realist, skeptical, even cynical view of human nature that animates Hobbes's Leviathan. So I'll probably use more breakout groups even in small classes then I have used in the past. The January term, again, I think it's a wonderful thing that the college is doing. We've reduced the standard course load from four to three and allow students to take a fourth course if you will, in January.

Austin Sarat (00:57:31):
And those classes again are going to be entirely online and there'll be very intensive. There'll be very intensive classes. And by the way, I should say that all students are only now required to take three courses. Many of our students are going to take the full four-course load that they would otherwise have taken. What matters more than it ever has mattered is the quality of advising that students get. And again, we have a wonderful program called the intensive advising program that pairs a couple of students, this year, it'll be three students with a faculty member, and that faculty member meets with them every two weeks. So advising at Amherst is no longer just about course registration and that kind of advising, which we began to develop before we went into the virtual world.

Austin Sarat (00:58:29):
I think we've seen how effective it is in the virtual world, and we're going to continue it and hopefully expand it when we get back on campus. And like my students I have to say this, I cannot wait to be back on campus. I'm suffering a little bit of heartbreak. One of my favorite Amherst events is move-in day. So for the last several years, Amherst's organized this absolutely wonderful move-in day, where when students arrive, they come to the quad and they're greeted by students and staff and faculty who are there to swarm the cars and help them move into their rooms. And I've loved doing that. I love schlepping the bags of students up into their dorm rooms and seeing the excitement as students greet their roommates. Well, we're not doing that this year.

Austin Sarat (00:59:30):
And I'll miss it, but I look forward to doing it. I look forward to doing it next year. Last year when I did it, it seemed that every student I moved in was on the fourth floor. And I think I said towards the end of my shift, I was exhausted from schlepping everything to the fourth floor. And I think I asked one of the student affairs people, like, was it just me? I mean, it was the luck of the draw. And she said, well, you want to be like in the first shift and we do it like some airlines, you know, by floor four, three, two, one. So she said next year, come back closer to the noon and we'll get you some first-floor assignments. So next year I'm looking to go back for move-in day about noon and get the first floor assignments.

Betsy Cannon Smith (01:00:13):
Well, there you go. I think you're signing up early enough. Well, Austin, this has been a real treat. You've given everyone who's on with us tonight, their own little sense of orientation. So no balloons, no swarming faculty, no fancy t-shirts, but I think you did the best we absolutely could. And this is an exciting semester we're heading into. It's so wonderful to have people like you and all of your colleagues who've spent so, so many hours making this possible for this to happen, starting August 24th on this campus. We appreciate you. I appreciate you. So a lot of fun to talk with you. Thank you.

Austin Sarat (01:00:48):
Wait, wait, wait. One more thing. So I want people to realize, people who are listening that what we are doing is more important than ever. The world needs the talent, the creativity, and dare I say the irreverence of our students. The educational mission of the liberal arts is more important than it has ever been. And I feel that in my bones. Wait till students arrive. This college is so well run, so well managed. So thoughtful in what it has done. When students arrive, they're going to come to a campus which is excited to have them, and students who are learning virtually, they're also going to be connecting to classes where faculty are excited to work with them. So we need people's excitement. We need their commitment. We need their belief that what we are doing is more important than, it's more important than ever. And I want to thank you, Betsy, and thanks to people who were listening and thank your colleagues behind the scenes who arranged this, the miraculous Austin Huot, who runs Conferences and Special Events. Thank you for making it happen and good to see you and stay safe.

Betsy Cannon Smith (01:02:20):
Thank you, Austin. And on that note, we will be having another one of these wonderful conversations. So mark your calendars. We'll send you details, but September 10, 12:15 eastern time for a conversation between ' class of "07 who's co-anchor of Early Start, you can catch her at 5:00 AM on CNN, and the James J. Gross Field Professor and Chair of Law, Jurisprudence, and Social, Lawrence Douglas for a conversation on his book, Will He Go?: Trump and the Looming Election Meltdown in 2020. So please join us for that. We'll be having a series of these throughout the year and look forward to seeing more. Austin, goodnight. Thank you. Goodnight everyone. Thank you for being with us.

Nick Horton (00:06):
Welcome to everyone to Amherst Conversations. Thank you very much for joining us for today's session. My name is Nick Horton, Beitzel Professor in Technology and Society. I'm pleased to be your moderator for what promises to be a stimulating hour. Let me start with some housekeeping issues. The event will be recorded. So it could be shared with those who are unable to attend. Once the discussion begins, the Q&A feature will be opened up for live submissions. You can locate the Q&A box next to this live stream feed. We live in a world full of data with systems that automatically extract meaning from it, and often make decisions on our behalf. These systems are increasingly autonomous and are making higher stakes decisions every day, but there are many important questions that we need to consider. How do we ensure that these systems are robust and reflect best judgements?

Nick Horton (01:04):
How can we prevent bad actors from influencing these decisions in malicious or unexpected ways? The type of machine learning models that Scott will talk about today are familiar to all of us. Imagine, for example, a spam filter for your email that classifies messages as spam, unwanted unsolicited communications or valid messages. When spam first became an issue, simple filters were developed that worked fairly effectively to classify messages in some cases, by using certain words as keys to let the message pass through. But then the spammers started to retaliate. They started to inject those same keywords into subject lines to trick the classifier. More spam started to get through. Here the consequences eally aren't so severe. If the scammer can game the system. You know, it's relatively modest if there's another unwanted email message in one's inbox, but other examples have far worse implications. In 2019, security researchers at McAfee added two strips of tape to a speed limit sign.

Nick Horton (02:16):
And were able to fool the autonomous driving system in a 2016 Tesla into driving 50 miles per hour over the speed limit. What other weaknesses do these self-driving systems have that could be exploited given the increasingly sophisticated and powerful systems that are now being created and deployed, there's a really important need to rigorously test and ensure that these systems are reliable. I'm pleased to report that these are questions really at the core of a liberal arts education that are being studied by students and faculty at Amherst College. Today's talk and discussion will address the topic of adversarial machine learning, and how faculty students at the college are contributing to both the science and the practice of automated decision-making. Our speaker today is Scott Alfeld, assistant professor of computer science, who teaches topics in AI, machine learning and security ranging from the highly practical to the purely theoretical.

Nick Horton (03:21):
He graduated from the University of Utah and earned both an MS and PhD from the department of computer sciences at the University of Wisconsin, Madison. Alfeld's primary research is at the intersection of machine learning and security. They study settings where an intelligent adversary has limited access to perturbed data that's fed into a learned or learning system. The goal of the research is twofold: to both detect attacks and to build an augment learning systems, to be more robust, to undetected attacks. In addition, they develop methods for inferring properties of the underlying sensors, whether trustworthy or not, and incorporating that knowledge into the data analysis pipeline. Before coming to Amherst in 2017, Scott taught computer science and public speaking and debate professionally in Salt Lake City, Utah, and Madison, Wisconsin. As a volunteer, they gave guest lectures for courses from the Wisconsin Center for Academically Talented Youth, and taught lock sport, recreational lock-picking and related physical security topics through sector 67 in Madison. Just a quick reminder that once the discussion begins, the Q&A feature will be opened up for live submissions. Attendees can locate the Q&A box next to the live stream feed. Scott, thanks again for agreeing to share your wisdom with the community. I'm really looking forward to hearing from you and later moderating the discussion.

Scott Alfeld (04:53):
Thank you, Nick. Thank you all for coming. I want to talk first about what we do as adversarial learning researchers, and then I'll go into some specifics of what my students have done. So just as Nick said, machine learning algorithms are out there and they make decisions and those decisions affect people.

Scott Alfeld (05:12):
So they affect people like Alice here. Now, as Nick mentioned, a lot of these, they vaguely change Alice's life, right.

Scott Alfeld (05:20):
What gets recommended on Amazon or, you know, credit card transactions being fraudulent or not many decisions are made without any human being involved, but more and more decisions that really affect a person are being made. So things like should this person be denied alone, or what should we set this person's bail to be come from these automated decision making systems with no human involved. And as Nick alluded to with the self driving cars, this is becoming more and more of a problem, right? We are putting engines behind AI that can then, you know, go 50 miles an hour over the limit or whatnot.

Scott Alfeld (05:58):
Now people try to trick these systems. So innocent little Alice here is a hacker. And what she might do is manipulate the data, fed into these systems, so as to trick them into doing what benefits her. We'll see a variety of examples of this. I'll mention this is separate from the sort of traditional hacking, which is more on the system security side of things. So what we do in AML is look at how do we understand these forms of attack and how do we make defenses against them? So if you get nothing else out of this talk, just know that my students and I are making life harder for the bad guys. In addition to that, there are some research aspects to this that by understanding the vulnerability of these learning systems, we're able to understand the learning systems themselves. So I'm going to talk a bit about just terminology so that we're all on the same page where how this machine learning will work is you give me what we're going to call a training set.

Scott Alfeld (06:54):
So you might say, okay, here are a whole bunch of medical records where for each one I've measured a bunch of things about the patients, right? That what's their blood pressure, what's their whatnot. And it comes with a label saying this patient either responded well or responded poorly to say some medication. So I, the machine learner, I take this training such and I produce what's called a model. For our purposes, a model is just a computer program, right? So I'm a machine learner. You give me a bunch of data. I give you a program that you can run. And then what you do, is you use that model to make a prediction. So you say, hey, this new person has come in. I'm going to run all these tests on them. Now I have these features. I don't know if they're going to respond well or poorly to this medication.

Scott Alfeld (07:42):
I haven't given it to them yet. But when I plug all that data into this model, it tells me its best guess as to whether or not it performs, they'll respond well. Now this is a very general setup. It doesn't need to be medical records. It could be something like you give me historical records of say, stock prices. So you say, this is the price on Monday, this is the price on Tuesday, Wednesday, Thursday, and so on. Again, this machine learning produces a model, just a computer program. And then you can go to that program and say, okay, I'm typing in the last 10 days of stock prices. Give me your best guess as to what tomorrow's stock price will be. So I'm going to stick with these two examples, but of course there are many others where machine learning gets involved. So what I'm going to talk about today all is based on this particular pipeline.

Scott Alfeld (08:34):
So I've got some same medical record and some label for that medical record. And then I have a bunch of these and together they make the training set. This has what were called feature vectors of all the things you measured. And then the labels saying, this is what I'd like to predict in the future responds well or responds poorly. This gets fed into a machine learner. Mathematical magic happens and out pops a model. Then we say, okay, great. I've got this model. I've got this new say medical record where I don't know the label, but I have all the measurements, all the features, we'll call this a test point. We pass that into the model, and what it passes out is a prediction. So this prediction says either, I think this person will respond well to the medication, or I think they will respond poorly.

Scott Alfeld (09:30):
This top bit, we're going to call training time. And then this bottom bit we'll call either test time or deployment time. Just keep in mind, there's sort of the training part where it says, here's a whole bunch of data where we know the answers, go do your magic to produce a model. And then there's the deployment part where it says, okay, you're done training, now you've deployed a model. It's fixed, just make predictions. And importantly, remember that this machine learner is a machine, right? This is a computer doing things. It's not a human. So it's, you know, some piece of code is doing this. So in today's talk, I'm going to talk about three projects. The first looks at defense. So we're going to say, suppose I have some model I'm worried about it being attacked. I'm worried about someone manipulating data. So how can I make it more robust. Then I'm going to look at attack.

Scott Alfeld (10:19):
So here it says, if I have some limited access to the training data, if I can make perturbations to it, how can I bend the machine learner to my will? And then we'll look at privacy issues, which says, okay, there's some training set out there. I want to know things about it. You've tried to keep it hidden, but through giving me limited access to your model, maybe I can reverse engineer aspects of that. So, I know that the name of the talk is Attack, Defense, Steal, that had a better ring to it. This makes more sense, pedagogically. So it's the order we're going. So first, this is work I did with Mackenzie. So she did her senior thesis with me. And then shortly afterwards, we turned that into a paper. This is currently under review. If you remember our pipeline here, what Mackenzie's work does is it says, we're going to assume that our attacker is affecting the test point.

Scott Alfeld (11:17):
So imagine that our attacker here is say a doctor who gets paid by some pharmaceutical company. And they say, Oh, I really want my patients to be flagged as responds well to this medication, right? The more people that I'm able to prescribe this medication to, the more money I get. So I'm going to perturb these test points, right? I'm going to perturb the records. Now presumably I can't, you know, say, Oh, this two year old is pregnant, right? I don't have complete control over how I change these medical records. But maybe I can lie about the weight of the patient, a few pounds or something just to change it in my favor. So Mackenzie's work does, is she says, we're going to take the machine learning algorithm and we're going to augment it so that it knows that attacks are coming. It says, when I deploy a model, not all of my data will be clean, some of it will be attacked.

Scott Alfeld (12:11):
So instead, I'm going to learn a more robust model. And this more robust model will say still perform reasonably well. We still want it to make correct predictions as much as possible, but it will be resistant to these fluctuations and data caused by an adversary. So, I'm just going to go through some of the highlights of this project. Back three years ago, some schmuck figured out how to do this optimal defense when you have a very small set of actions. So you can imagine, you have your machine learner and then you get to pick one of like five things. So these things could be, you know, I have five different models to pick from, or I have five different ways of making it harder to attack or whatnot. This work said, okay, here's how you figure out the best one to do. What Mackenzie did is she took that and extended it to a setting where now I have a continuum.

Scott Alfeld (13:10):
So I have an infinite set of possible actions, and that set can be continuous. What this did is it opened the door for saying, well, Hey, my actions can be to change my model to any other model. And so now it says, let me find a model that is robust to attacks. Now it's pretty easy to make a model robust to a text. All you do is say, ignore the data entirely and just return responds well. That's not a great model, right? And there's this sacrifice and accuracy for a gain in robustness. So through a whole bunch of experiments, we found it's worth it. You can get a huge reduction in the vulnerability for a very small reduction in the accuracy. And what's more with Mackenzie's framework, you can tune it depending on what you want. So if you say, well, look, I mean this data, you know, I just asked a thousand people for it.

Speaker 2 (14:05):
And every single one of them has an incentive to lie. I'm really worried about it being attacked. You just turn one knob and Mackenzie's algorithm, and it says, okay, you're, you're more robust and you'll sacrifice some accuracy for it. If you're doing weather prediction and you're collecting your own data and you say, I don't think there's any adversary out there, then fine. You turn that knob the other way. You're now balancing more for accuracy, less for robustness. Now, in order to understand how to defend, we need to understand the attacks and how it works. So as a separate project, this is work I did with Lucas. He also did his senior thesis with me at the same time. And then this led to a publication in AAAI 2019. AAAI is one of the big AI conferences out there. So remember I said, we've got this pipeline.

Scott Alfeld (14:58):
I kind of lied to you. The pipeline is typically more complicated than this. So often what'll happen is there'll be some set of preprocessing steps. So when you get your data, maybe you synthesize some features from it. Maybe you clean it up, maybe you remove outliers. Maybe you normalize the features. You do some form of preprocessing. So you have this original data, it gets preprocessed, and then that turns into your training set, and then come test time. You have some unseen point. You do that same preprocessing and then it becomes your test point. Now, when Lucas started, the bulk of the work out there said the attacker, if they're acting at training time, affects the training set directly, right? So they say, okay, you've preprocessed your data, now I'm going to go in with my hand and muck about what that data. Now what Lucas, his work said let's go before that step.

Speaker 2 (15:54):
So you, as the attacker, you affected the original data, then it gets fed into the preprocessing, and from that point on you don't get to touch it. So in this work, Lucas here is behaving as the attacker. So I'm also gonna show some highlights of his work. One is that his work had this plug and play aspect to it. And what I mean by that is if we look at our pipeline, okay, I'm sorry. I lied again. Typically this preprocessing step is actually several different preprocessing steps. And what he did is say, okay, if you figured out how to attack A, and you figured out how to attack B, and you figured out how to attack C, and you figured out how to attack the learner in the old form of attack, right, where you can affect the training set directly. Well, then you figured out how to attack something that does A then B then C and then that learner, right?

Scott Alfeld (16:45):
You've also figured out how to attack something that just does B and then A and then C or whatever, right? By learning the little pieces of this pipeline and how to attack those, you can just click them together like Legos to figure out the best attack against the entire pipeline. In addition to this plug in play framework, this idea of preprocessing is actually very broad. So you might think of this as, Oh, okay. I'm going to remove some outliers. Maybe if I have my features, I'm going to scale them to be between zero and one, or maybe if I've measured a whole bunch of things, I'm going to convert them all to be, you know, metric units are all to be freedom units or whatever. It can be much more than that. So I'm going to show what's called the Hankel Transform, is an illustration of how these preprocessing steps can actually capture a lot.

Scott Alfeld (17:36):
So suppose we have stock price data. So here, I'm going to say we have some values. What was the price on Monday, Tuesday, Wednesday, Thursday, Friday, and Saturday? And now we want to build a system that says, given the few days, predict the next day. Well, we can phrase this as a data transformation step. So we'll say, we'll construct this matrix, it's called a Hankel Matrix, where we put the Monday here, the Tuesday gets this diagonal line, Wednesday gets this diagonal line, and so on. Then we take that matrix. We're going to chop off the first two columns.

Scott Alfeld (18:15):
So we've got those, and then chop off the last one. And now we say, okay, these are our features, and these are our targets. So if you look at each row across this, it says, well, first use Monday and Tuesday, and try to predict Wednesday. Then use Tuesday and Wednesday and try to predict Thursday. Then take Wednesday, Thursday, and try to predict Friday and so on. So this process of going from the sort of simple, I have a bunch of data points let me learn some predictor, to I have a time series and I want to learn a predictor. He was able to put into his framework as a simple preprocessing step. And so this allowed us to find attacks against these forecasters, right, these time series forecasters. In addition to various other preprocessing steps, what he also did and this made it into his thesis, but it was worked on after the paper is take limitations that the attacker would have.

Scott Alfeld (19:18):
So if you're familiar with channel stuffing attacks, it's this trick where basically I can say, okay, the quarter is almost over and I'm going to get charged a bunch of money based on how much inventory I have. So I'm not going to have any inventory. I'm going to put it all on trucks and send them on a big loop, you know, a hundred miles. So that at the moment that we switched to the new quarter, I have very low inventory, so I pay very low. So various forms of constraints that exist in the real world for what types of attacks you can do, he was able to put that into this same preprocessing framework to say, okay, now we can figure out what real attackers out there are actually capable of doing, which is the first step to figuring out how to defend against that.

Scott Alfeld (20:04):
And lastly, I'm going to talk about some work with Billy. So, Billy also did a senior thesis with me. This was the following year, and he presented some of this work at the AAAI 2020 Undergraduate Consortium. So we've got this pipeline here. I'm going to ignore a couple bits of it. So for this work, ignore the preprocessing step and we can fit it in later, if we want to. And for now ignore actually using your model. So, you know, okay, you've got these test points in these predictions, forget about that. Suppose we just have this framework. So you can imagine this as a setting where Amazon does this, for example, they have a huge amount of data. They pump it through some machine learning, they have a model, and then they give you access to that model, right? They say, Hey, if you give us points, we'll predict them for you.

Speaker 2 (20:53):
Your phone probably does this, too. You give it an image and it highlights where the face is, right? It's simply saying we did the training some time ago. Here's access to the model. Now, of course, I kind of lied to you. This isn't really how the pipeline works. In practice, what would happen is, time would pass and you would get new data. You would say, Oh, Hey look, I've collected new medical records. And I gave them the drug and I've seen whether or not they performed well or not. Right? And then this new data you say, well, look, I have this new data. I might as well add it to my training set. Now I've got this bigger training set. Let me pump that into the machine learner. That'll crank out some new model, and now I have this new model. Then, maybe I'll go ahead and deploy that model.

Scott Alfeld (21:36):
I'll say, Oh, okay, I have this new patient here. Let me, you know, get a prediction. Well, of course time goes on and this continues. So I'd get some additional data. I now have this newer data. So I'm going to take that, put it into my training, set, pump that into the machine learner and get an even newer model. And this process continues. So in Billy's work, he says, okay, if I get to see the newer model, so say I, you know, query these Amazon models and I learn what they're doing every so often over time, I see the updated models. And suppose I see something about the new data, but importantly, I don't see anything about the original training set, but I want to. What Billy's work does is it says, okay, by observing how this learner changes over time or how the learned model changes over time, with this updated training set, what can I infer about the original training set?

Scott Alfeld (22:33):
So this, and I'll mention here, Billy was being the attacker in this. This led to some really cool math. I'm not going to go into the details of it, but what it captured, there were some impossibility results saying for this particular learner, under a bunch of assumptions, there are things you fundamentally cannot learn about the training set. And you can think of this is a whole bunch of machine learning says, Oh, what does it learn? Right when we give it a bunch of data and it does all this math and it produces this model, what did it actually learn? Billy's work answers the question: what did it forget? What aspects of this training set were lost in the process? In addition to that, privacy is really important. So there are security and modifications of privacy. You can imagine if, let's call it the President of Enemyland shoots down an American drone.

Scott Alfeld (23:28):
While on this drone, there might be a learned model that says, you know, given an image classified as military compound or not military compound. But the training data, of course, is not kept on the drone. But if I'm the President of Enemyland, and I have access to this model, I might say, my number two has this aerial photograph. They're the only person in the world who has this aerial photograph. And I really, really want to know if this photograph was used in the training set for this model. If it was, my number two is an unlucky guy, no longer my number two. Right? So insecurity domains, you want to know by releasing access to the model, what are you actually revealing about the training site. Similarly in the medical domain? So you could imagine a setting where I say, Hey, potential employee, thanks for coming and interviewing here.

Scott Alfeld (24:22):
Now I know that you participated in this medical study and I know certain things about you, right? I know your height, your age and whatnot. I really want to know whether or not you have diabetes, say it's a hobby I have. Can I look at some model that was trained on your data as well as other people's data and reverse engineer certain aspects of your medical record in that training set. Right? Typically we want to design our machine learning setup such that if I give you access to the model, I'm not revealing any private information about the elements of my training set. And of course there are many other reasons that people care about privacy. And hopefully I don't need to convince you that privacy is important. I'm going to quickly cover a few things I've done recently, not related to students.

Scott Alfeld (25:15):
Some of the stuff I've looked at is what if there are multiple attackers at play, they might be aligned. They might be competing with one another. And then similarly, what if there are multiple defenders? So you can imagine a setting where I have some message that I would like to send, and I would like to beat your spam filter. I don't really care about your spam filter. I just want to send it to a million people and bypass as many spam filters as I can. So, you know, I'm sending this message out and there are a thousand defenders. I don't care which ones I beat, I I just want to construct the attack to beat as many as possible. And then similarly, these thousand defenders, they know that and they want to tune their systems such that they can best defend in this competitive environment. Then I've also done a bunch of stuff looking at graphs. So I say, if, you know, the attackers are attacking a network, or if agents are interacting on a network where what happens to me affects you in some way, how can we best understand how to defend these, these more complicated systems where the machine learner is just an integrated part of it all.

Scott Alfeld (26:26):
I do other things as well. I'm going to skip most of them. So I do want to thank some people first and foremost, of course my students, this talk would have been incredibly boring without their contributions. I'll mention Mackenzie is now a software design engineer over at Amazon. Lucas is now in Indiana pursuing a master of music and vocal performance. And then Billy is also at Amazon. Especially on the work with Lucas and Billy, I want to thank some collaborators, namely Ara and Ben. And then as far as funding goes, so the summer after graduating, Mackenzie was awarded this fellowship that allowed her to come back for a few weeks over the summer, and that's when we wrote the paper. So thank you to all of you donors present who made that possible. And I want to give a special shout out to Alex and Hillary. I don't know if they're here or not, but if so, hi. They all of these students benefited from the McGeoch Fellowships over the years. And with that, I will open it up to questions.

Nick Horton (27:36):
Great, Scott. Well, a really interesting, fascinating, and at some level, quite disturbing talk in terms of the need to be thinking about these questions of attack, defend and steal, and the many applications of these methods are coming in. I was struck during your presentation about how it really brings together a number of areas, you referred a couple of times to the underlying math and kind of probability that's part of differential privacy, kind of big questions that are kind of now out there in terms of data science. The kind of game theoretic aspect of this also seems, seems really, really relevant in so many ways. And I guess my question would be coming back to Mackenzie's project, how do you balance the different parts and pieces here? You gave us one example where, you know, you were assuming that it was just all bad actors providing data to you, and another example where, you know, you were saying there was not, but, but in the middle, how do you, how does one make decisions and some of those areas?

Scott Alfeld (28:41):
Yeah. So there's this issue of I've been given training data, and then I want to learn a model that predicts future training data. So in essence, I'm learning how the world works, right?

Scott Alfeld (28:59):
The data arose from the world in some way, some phenomenon caused it to be the way it is. Now, if you're the attacker and suppose you have access to 1% of my training data, right? So you can perturb 1% of it. Well, then what I want to do is be robust to that. I want to say, no, no, no, your attacks, they're not going to change what I learned that much. Right? I'm learning the core of nature. Now, if you have access to a hundred percent of my data, well, then you are the world. And the only thing I can do is learn how you create this data, right? And so there's an underlying assumption often made whether implicit or explicit that is big world, small attacker, right? If the attacker is really, you know, if they're controlling sixty percent of the data, well, then who's to say that they're not nature, and that what we call nature before is actually, you know, some separate, random process.

Scott Alfeld (29:51):
What Mackenzie's work did in particular is that 1% becomes tuneable, right? So do we, it's not exactly affecting 1% of the data, but, you know, put a bound on how much you can perturb points, say, on medical records, if you lie about a patient's weight by plus or minus five pounds, that's fine, but you lie by 150 pounds, that's not OK. She's able to tune her system based on how afraid you are of the attacker's strength, but it's always under the assumption that that attacker strength is fairly small.

Nick Horton (30:28):
Great. So some questions are coming in now. Here's one, do two hackers actually perform data manipulation attacks?

Scott Alfeld (30:37):
That's a great question. Yes, we don't know, or at least, I don't know the extent of it. So a simple example of this is suppose I write a book and I sell that book on Amazon and I wanted to sell more.

Scott Alfeld (30:55):
Well, what I can do is I can create a fake Amazon account and then go review that book and give it a very positive review. Right now, if I'm clever, I can go write a script that creates 10,000 Amazon accounts, and then strategically browses Amazon, ending up landing on that book to change what it learns. Right? So this sort of you know, astroturfing, as an example of this attack. Now, how sophisticated these attacks are, there are certainly cases where they're pretty sophisticated. There was a company I'm not gonna say their name, that would do this as a service. So if you had an app that you wanted to sell on the iOS market or the Android market, you could pay this company. And what they would do is they would take your app and attack the market for you so that it would learn to recommend your app more. So the attacks certainly do happen in any situation where there's money involved. So in, you know, stock trading or Amazon sales or whatnot, I'm sure a lot of these attacks are happening. The level of sophistication behind them varies a lot. And it's not really known where the, how sophisticated most of the attacks are.

Nick Horton (32:16):
Great. There's a kind of a question about, about these models and kind of constraints on them. Cause you talked about kind of like you could see the model or you could query it in some way. You know, what kind of results would hold if the learner really had unfettered access to kind of getting a computable function from that model, is that, is that a kind of an interesting scenario and is that kind of realistic in some of these settings?

Scott Alfeld (32:47):
Yeah. So the, how a learner works is it says, this is the space of all possible models I might end up learning. And maybe this is really a simple collection, right, it just says, Oh, okay, whatever, you know, my prediction for tomorrow's stock price is going to be some linear combination of the previous five days, or maybe it's something far more complicated.

Scott Alfeld (33:11):
And then what it does is it selects some model from this space. And there's a question of, if I'm learning a super simple model, it's conceivable that I'm more robust, right? Because you can't make me learn something that's all that complicated. And so if your goal is that, you know, I predict everything perfectly, except for when I see this particular pattern, at which point I predict that it will be a trillion dollars. I simply as the learner don't have the capability of doing that, these extremely complicated models on the other hand, like what's behind self driving cars. They can learn that they can learn all sorts of weird things. And what people have seen in the community is those attacks are super duper susceptible to these forms of attacks. So as you mentioned, Nick, you know, you can put a couple pieces of tape on a stop sign.

Scott Alfeld (34:04):
And I mentioned, you place the tape very strategically on this stop sign. And it tricks a self driving car to see the stop sign misinterpret, the, you know, it takes an image of it, and then it passes that image through a learned model, misinterpreting it to a speed limit 2,000 miles per hour or whatnot sign. So as you get these more and more complicated learners making these more and more complicated models, the susceptibility to attack potentially goes up and it becomes...the attacker in some sense has more power in what they might be able to achieve.

Nick Horton (34:48):
Great. So fairness, transparency and accountability have become more prominent conversations in the machine learning world. We could talk a little bit more about how this plays out in the curriculum at the college later on. But the question from the audience: can machine learning take human bias out of gateway pipeline processes in our society?

Nick Horton (35:12):
And they gave examples, admissions hiring promotion. And then how do you see those aspects playing out in terms of the adversarial modeling work that you're doing?

Scott Alfeld (35:24):
That's a great question. So the, let me make sure we're all on the same page and giving an example of fairness. When people talk about algorithmic fairness or fairness in the context of ML, suppose I have a model that predicts, you know, responds well or responds poorly to this medication and it's 95% accurate. So 95% of the time I correctly guess, and then you have a model which is 94% accurate. Well, you might just say, Oh, Hey, let's go with the 95. That's better. Well, if we drill down into it and we find, say, there are two groups of people, we'll just call them majority and minority, and I get 99% accuracy on the majority and 50 on the minority and you get 94 on the majority and 94 on the minority, in some sense you're more fair, right?

Scott Alfeld (36:18):
Yes. Technically across the entire population I perform better on, you know, on my model is more accurate, but you're more fair. So this is the issue of fairness. And I'll mention it's complicated. We can't just have a system where, Oh, if you're in the majority, you get plugged into my machine learning model, and if you're in the minority, you get plugged into this other one and then everybody's happy. As you can imagine, this is problematic and that you go to a doctor and they say, Oh, yes, here it will test you step through this door. Your friend goes through that door. Trust me, it's just as good, we totally put just as much funding behind that door. So this is the issue of fairness. What adversarial learning does is let us better understand how learners respond to changes in a training set.

Scott Alfeld (37:08):
Right? We do this through the lens of an attacker who says, Oh, I'm trying to change it in this particular way. That benefits me. You can directly apply the methods by saying I'm an attacker. I want your learned model to be more fair. How can I perturb the data in such a way that it becomes, what you learn is more fair? You can also say, all right, well, through our understanding of adversarial learning, how attacks work how we're able to defend against them, can we craft a model or a learner that is more robust to bias being put in the training data. If we can, then that's great, right. And we'll say, okay, this is now more robust in a different sense, right, it's more fair, separate from that on the interoperability side, the adversarial machine learning for mathematical reasons connects fairly directly to interpretability of models.

Scott Alfeld (38:10):
So right now, if I have my self driving car and it sees a stop sign and it says, Hey, that's a 3,000 mile an hour sign. I can't say to the car. Why, why did you, this is clearly a stop sign, what, what, what were you doing here? Or at least if I do say that I get a very hard to interpret answer back. And some of the methods from adversarial learning have helped us ask those and say, can we make this model capable of justifying its decision or at least explaining its decisions? And if so, that'll help us, you know, audit these systems when it comes to issues of, of bias or fairness.

Nick Horton (38:50):
Yeah. There's kind of a follow up question to that. That analogy with the the tape on the traffic signal, and it's kind of thinking about medical databases and you talked a little bit about one of those, those examples. You know, is it usual, or can the training set have hard stops on determining kind of like these features or artifacts, like that your heart rate has to be below something, or as you said that the weights are within some kind of usual ranges, is there any ways to automate that process of ensuring that the training data has those, is that what the defense entails?

Scott Alfeld (39:29):
That's a great question. So the broadly speaking, say, I want to write a paper on some attack, right? I have an idea for, oh, this, this learning method is susceptible to this attack. So I'm going to formulate what the attacker can do, what the defender does, and then I go off and I write a paper for it. Well, what I need to do is find a description of the attacker, which has two aspects to it. One is that it needs to relate at least vaguely—I mean, I am an academic—to reality, and second, it needs to be mathematically friendly enough to where I can actually say, okay, now we can solve the optimization problem. Right? We can figure out what the the best attack is. The hard limits on saying a heart rate, can't be above blah, you know, the various aspects have these sort of are hard cutoffs and strange aspects to them.

Scott Alfeld (40:24):
You know, they might relate to each other in certain ways that makes it very hard to end up with a mathematically friendly problem. So there are heuristic ways of taking a guess and saying, yeah, this attack may not be the best, but it's probably pretty good. And then empirically, we verify that it works well. And on the defense side, there are certainly times where you can say, so I did some work for a company years ago doing inhuman traffic detection. So this was, if you have a webpage and you get traffic to it, you want to know how much of that was from humans and how much of that was from the bots that roam the internet. As a first pass, you can do a lot of things like, Oh, Hey, look, this user clicked on this link a hundred thousand times in a 10th of a second—probably a bot. This, you know, various what you see in the impression of the traffic, you say, Oh, this value is negative.

Scott Alfeld (41:21):
That's clearly about or whatnot. That's a good first pass. But then there are the, you know, clever attacks that are constraining themselves to the valid values, right? The valid inputs to the system. And so that definitely works as a first pass of a defense system, but then you typically need something more beyond that to catch the trickier attackers.

Nick Horton (41:47):
Great. So a more technical question for you: can one aggregate the attack transformations at different steps of the preprocessing pipeline with the same results? Is it, you talked a little bit about separating those are kind of mixing matching, I believe there was a Lego metaphor involved.

Scott Alfeld (42:08):
Yeah. So so, so what was the question about this, is can I aggregate the...

Nick Horton (42:13):
Can you aggregate the, you know, the attack transformations at those different steps?

Scott Alfeld (42:23):
Okay. So how this works is, I have my data is some function of some function of some function of some function of the original data. I take my original data, I pass it through some data transformation preprocessing step. Then I take that, pass it through some other, and so I get this composition of functions. And what Lucas' framework leverages is the chain rule from calculus. So it says if I need to figure out how I affect the...you can kind of, here's a potentially reasonable way of thinking about it. I'm an attacker. I say, this is the attack I'm currently considering. Now I want to figure out how can I change it slightly. If I say, Oh, should I change it in this direction?

Scott Alfeld (43:11):
Well, I want to know how I, how that change affects the attack once it's passed through the entire pipeline. I can figure out how it affects it at each individual step, and then through the chain rule from calculus, that lets me determine how it is, how it affects the entire pipeline.

Nick Horton (43:34):
So just on behalf of my math and stat colleagues, I do appreciate the shout out to the chain rule. So that's the fact that you said it twice was more than we'd agreed to. Question about sample size does your work assume very large quantities of data in those, and are there times when the data set is small but critical, and what impact does that have on, on these three strategies?

Scott Alfeld (43:58):
That's a great question. So my work typically focuses on machine learning methods that work on relatively small amounts of data.

Speaker 2 (44:13):
So there are models out there, if you look at like deep learning models that are methods of, if you look at deep learning, it often takes a huge amount of data to produce a model, and they're able to benefit from that huge amount data and benefit from that huge amount of computational power to produce, you know these really amazing models. These are what's behind the self driving cars that we've been talking about. Much of what I do is closer to the the simpler models that behave on much smaller data sets and use less computation. One of the advantages of this is you're able to analyze them and get stronger guarantees often. It says, this is indeed the optimal attack, there is no attack better than this. And then some of the work I do also says the learner is just a component in this system.

Scott Alfeld (45:16):
It could be one of these simple ones that, you know, we have all these results for. It could be one of these extremely complicated ones and extremely computationally, expensive ones. Either way, what it can just be plugged into my framework and then go from there. But typically I work with the models that are utilizing less data.

Nick Horton (45:41):
Are certain machine learning algorithms, more robust against the kind of reverse engineering and adversarial issues that you've described and others?

Scott Alfeld (45:57):
Yeah, definitely. So the, you know, I mentioned this idea of my machine learner, you know, that's gonna predict tomorrow's stock price always says $15. This is an incredibly robust model. You can perturb the training set, however you want. You can perturb the test set however you want. It does not change, it's fully robust. On an extreme you have, you know, I'm classifying images for my self driving car.

Scott Alfeld (46:26):
If you perturb a few pixels by a little bit, I will dramatically change what I, my prediction is. There is a spectrum between these, right. You can have, Oh, okay, I'm, you know, learning a linear function. I'm in a low dimensional space. I am more, I am harder to trick. There's also a question of the attacker's intentions and the underlying data distribution. So if I'm trying to classify images between cat and dog, and you as the attacker, you know, you have a picture of, I dunno, let's say that I'm doing more fine grain detail of that. You have a picture of a cat and you want me to predict Chihuahua, maybe that's an easy task. You have a picture of a cat and you want me to predict great dane, maybe that's harder. So the robustness depends not just on the learner, but on the learned model and on the attacker's goals and capabilities.

Scott Alfeld (47:35):
So like in Mackenzie's work, what she did is said here is the space of all the models I might learn. How can I take what I would learn, assuming there's no attacker, and shift it to be a different model that has similar accuracy, but much higher robustness, much higher resistance to these attacks.

Nick Horton (47:55):
So you're starting your fourth year at the college. And the question comes up, what will you work on next? What's anything in future we should start to be looking for?

Scott Alfeld (48:06):
So I have changed my plans for this year a bit. I'm on sabbatical this coming year, and some trips got canceled. I think I'm going to do stuff looking at the— One way to think of this as adversarial AI. So there's this whole question of where machine learning and AI land, but basically there are methods of AI that look at less data-driven approaches to learning, perhaps more experience approaches to learning.

Scott Alfeld (48:42):
So like how can I learn to play chess, these sorts of things. And I'm very interested in how those can be attacked and defended and perturbed. So that aspect of it, I find very interesting and I hope to explore that more. Just as a hobby, I'm looking into quantum computing. I have some theories about how some aspects of quantum computing relate to the optimization problems we often see in adversarial learning, which I'm pursuing. And I have a thesis student this year, Sam, who I'm very excited to work with, and I'm sure she will do great work. And I will see where she, what path she leads me down in research.

Nick Horton (49:27):
So that's kind of a great transition to one of mine, which is what's it like teaching machine learning to Amherst students?

Scott Alfeld (49:35):
It's great. Like I like the students a lot.

Scott Alfeld (49:38):
And one of the so I've taught at other institutions too. And one of the things I've noticed here is the in part due to the open curriculum and the structure of prereq requirements and whatnot, in part due to other reasons, there's a huge breadth of background in any single class. So I'll have students who, you know, they're in their last semester of a math major. And they're interested in the machine learning side and they've never written a line of code. I have other students who, you know, have never seen any calculus before and are early on in their computer science career. And I need to design the course in such a way that they both get something out of it, which is terrifying, but also a lot of fun. And then in addition, whenever we bring in something outside the core of ML, which happens all the time in adversarial learning, students have these different perspectives.

Scott Alfeld (50:40):
So like I love discussing fairness because these students, you know, they say, Oh, in this thing that I'm interested in, this came up and here was the issue, and you know, what you talked about before and machine learning land can apply directly to that. And these making connections with very different fields is great. And it's also super helpful in that, you know, if you're taking the role of the attacker, well, you get to do anything you're capable of doing. And the students have been great about coming up with motivations behind attackers, figuring out what is realistic, what's possible, and at the same time, they've got the mathematical chops to, you know, run through it and, and do really, really great work. So in teaching the courses and advising students for thesis and for summer research and whatnot, it's been fantastic.

Nick Horton (51:35):
Great. A hypothetical question. Various things that have been coming in on the, on the Q&A. Imagine that you've been asked during your sabbatical, which you've earned, and we're going to excited to see what you do during that, but say that, you know, the president's science advisor pulls you in for a consult. And they're interested in trying to understand, you know, whether machine learning models are being in reinforcing red line red lining, whether there's the dark trace algorithm, they don't know what's going on with that algorithm. They've heard that, you know, that there's, there's credible evidence that that a foreign nation has been kind of stealing data from Anthem health systems. And they're trying to get their hands around this one. What advice would you give them advice?

Scott Alfeld (52:26):
That's interesting. So my go to advice whenever I get any security question is typically use two factor authentication that doesn't seem to apply in this setting. I think the, in my work at least, and I think generally more globally, there's a fairly solid line between detecting an attack and then handling it outside, as in Hey attacker, how about you and me go take this outside, right? Where you say, okay, I've detected that you've done this, so I will shut down your account. I will delete the data whatnot. And then there's a wall between that and I am building a system robust to attacks that I did not detect. Most of what I've done is in that box. I've done a little bit, so in this inhuman traffic detection stuff I talked about, I did the how to detect attacks. And, you know, a lot of it is to recognize everybody gets lucky.

Scott Alfeld (53:32):
So if you imagine I'm looking at the data and Oh, this person predicted stocks would go up and they did, and whatnot, everybody gets lucky. You want to look for the times that people are getting lucky more often than skill can explain how you do that varies dramatically based on the setup and, you know, depending on the details of what the president's science advisor has asked me about, I might have more concrete questions of or contributions to, but a big part of it is, you know, you're performing an investigation and there are many, many different angles you need to look at and get the sort of holistic view.

Nick Horton (54:14):
Great. Well, unfortunately, we're coming near the end of the time. Just one question left for you. What's this business in your bio about recreational lock picking? Have you been practicing during quarantine?

Scott Alfeld (54:34):
A bit. So for those of you don't know, lock sport is you take a lock, you try to pick it open usually quickly. And so, I've done a lot of lock sport in the past. I've done a lot of miscellaneous, you know, physical security stuff in the past, and I've connected that to computer security in my teaching. So I teach our computer security class, which is much more on the system security side of things, not this adversarial learning stuff so much, and many of the problems we face in computer security of, you know, how do I make this resistant to attack while still convenient? And how do I make it so I can handle multiple users and stuff? Many of these problems were solved decades ago for other systems where there was no electricity involved. So, you know, if you look at bank vault, safes, and various, you know, master keying methods and whatnot, there's a lot of there's a lot of value in saying, we're going to discuss this security topic, and we're going to begin with you seeing it in a piece of brass that you can hold in your hands and play with.

Scott Alfeld (55:50):
And so I've drifted from sort of the competition side of, you know, I want to be the best lock picker in the world to more of the, I want to teach people how to pick locks and how locks work and how physical security works in general. Near the start of the pandemic, I picked up safe cracking, which is a slow process to learn. But you know, now's a good time to do it.

Nick Horton (56:16):
Well, Scott, again, I'm glad that you're keeping yourself busy, that you're keeping yourself safe.

Nick Horton (56:24):
It's really remarkable the work that you're doing on behalf of our students and the college on a really important area that I do think really spans the curriculum. So I want to just thank you again for taking the time tonight to share those thoughts with with the alumni and our students and other members of the community. Thank you.

Scott Alfred (56:43):
Yeah. Thank you.