Graduate Teaching on Diversity: Week 8/9

Because of Thanksgiving, we only met in Week 8.  We had a guest lecture by Anne Gillies from our Office of Equity and Inclusion.  Anne gave an overview of implicit bias, what it is, how it arises and how we can overcome it.  It was the only full lecture we’ve had in the class so far, and it will be interesting to see what students prefer — I will be surveying students next week on various aspects of the course.  I think I would prefer, as a student, discussion-based classes, but students may be more comfortable, in a university setting, in a more standard lecture-based scenario.

But then this class is all about getting students out of their comfort zones.

Anne has attendees take some Implicit Association Tests before discussing implicit bias, which I find an interesting tactic.  Anne also came to our faculty retreat at the start of the quarter to give a too-short (not her choice) version of her talk.  There was some resistance from faculty to the notion of implicit bias and that the implicit association tests may say anything about one’s implicit biases. I had the same concerns when I first took these tests.  So, before the class, I asked Anne if she had some readings about just that, to preempt people’s scientific skepticism: this is a technical article and this article is less technical (and resulting Q&A).

There weren’t many questions, but of course, we didn’t have much time for questions (again, we need more time for these topics!).  I am reminded that, when inviting a guest lecturer, I need to do a better job of explaining the audience.  I fear that the majority international, many of whom are English-language learners, audience may have had difficulty with the pace of the lecture and the American cultural references.


I also looked over a short assignment to the class.  I had students read Never Meant to Survive: A Black Woman’s Journey: An Interview with Evelyn Hammonds, which my colleague Padma Akkaraju pointed me to.  It is an interview of a black women who studied physics and eventually left physics, despite academic success, in graduate school.  When I read this interview, I was struck by how many aspects of discrimination were described and yet, because it is an older article and an interview as opposed to a scholarly article, it doesn’t use a lot of the terminology that we use today (such as implicit bias and intersectionality).  So, as an assignment, I had students read the article and underline and annotate instances corresponding to a short vocabulary for diversity we studied in week 1.  Overall, there were quite thoughtful annotations and several students annotated similarities to their own experiences at OSU.  In future renditions of this course, I would put this assignment much earlier in the course — in week 2, for example.  I also quite liked this mechanism for an assigned reading, so will likely use it again (and recommend it for others).

Graduate Teaching on Diversity: Week 7

This week, I turned to dealing with incidences of ‘local’ discrimination.  I asked students to read these guidelines on how to challenge oppressive moments before class, with the goal of instilling the idea that it is not enough for one to not engage in racist or sexist behavior, but one must also respond to such acts on the behalf of the oppressed group.

I think something that is difficult for members of the dominant group to understand is that while any given racist or sexist incident can be written off (‘ignore it’, ‘treat the offender with the disdain they deserve’), for those in minority and underrepresented groups, these incidents can happen on a daily basis.  They build up to create an oppressive environment that is unwelcoming, at best, and threatening at worse.  In the university context, this environment supports the stereotypes that cause quantifiable reduction in academic performance. (I also asked students to read this overview of stereotype threat.)

At OSU on Monday, we were reminded of the impact of constant, background oppressive moments during a student-organized speak out for students of color.  Students bravely spoke in front of a mostly white audience of 500 about their daily lived experience on campus and how the daily micro-aggressions rise to make them not feel safe and invoke ideas of arming oneself in self-defense.  Unfortunately, but not surprising, at least to me, the daily acts of racism were not limited to coming from fellow students.  One student spoke about an instructor identifying black people by the n-word; another identified undocumented workers as illegal people and yet another, in teaching a Spanish language class, led a discussion that justified Cortés’ conquering of Mexico.

So, last Friday, I had my graduate students break into small groups of ~5.  Each group was given a scenario describing a real incident of sexism or racism that has happened at OSU or another institution that I have been at. I will list some of these incidents here that are sufficiently deanonymized, already publicly known, or that happened to me; that is, those that I feel I have the right to publish online:

  • A female student raises her hand in a tutorial (recitation) session and asks a question. A TA, instead of answering the student’s question says “Little girl, you will have to work a lot harder to keep up with the men in the class.”
  • A small group of graduate students are talking informally about their job prospects. One student says to a female student “well, you don’t have to worry, you’ll get a job through affirmative action.”
  • Graffiti has appeared in multiple bathrooms in KEC. The graffiti encourages violence against minorities and uses racial slurs.
  • In a small group with several graduate students and one faculty member, the faculty member indicates that the black student in the group was admitted because he was black.

Each group was asked to brainstorm ways in which they would respond to these situations if they were, for example, a fellow student, a TA, or simply a member of the community.  After 15-20 minutes of small group discussion, we went from group to group, sharing the incidence each group discussed and their ideas on responses.

IMG_20151113_165717 IMG_20151113_165724

There was some thoughtful discussion on when is the best time to respond, with some (possibly often valid) concern that responding in the moment could, in certain situations, make matters worse.  We also discussed, in the graffiti example, whether one should draw attention to it, so that people know this kind of thing happens, or if one should quietly remove it, to minimize the damage it does.  As with many things in the class, there isn’t a yes or no answer.  (Teaching algorithms is so much easier.)  I did, however, encourage students to report incidents, no matter how small they think they are, to the Office of Equity and Inclusion so that they can help determine how serious the matter is and whether it belongs to a pattern of behavior on campus.

I highly recommend this kind of activity.  I think it is very helpful to (a) hear about bigoted moments (b) imagine what it would be like for these kinds of things to happen on a daily basis and (c) prepare oneself to respond as an ally.  In the incidences of sexism that I have suffered, I felt utterly alone as those around me failed to respond.  I am hoping that with a little preparation, our graduate students will have the confidence to speak out against bigotry.  On the other side, I have also felt utterly unprepared, in the past, to respond to sexist comments.  I have had one student request that we have a short meeting for women in EECS to workshop how they could respond to sexist comments and behavior, which hopefully will happen next quarter.  OSU offers a retreat called Racial Aikido that helps students of color learn how to respond to racism that was referred to positively on Monday’s speak out session.  It is interesting that students of color are encouraged to spend two days learning how to respond to racism and other students are not expected to spend any time learning to not dish it out.

 

Graduate Teaching on Diversity: Week 6

Today, we held a discussion specifically on (overt) sexism in sciences.  I assigned three readings on recent-ish occurrences describing Nobel-Prize winner Tim Hunt’s comments, Satya Nadella’s advice to women to not ask for raises and Larry Sommer’s views on women’s ability in STEM fields.

I used a spokes council discussion, as I described two weeks ago, to have students work on the following task:

You have come together to decide if and how to respond to sexist statements made by notable academic, industry and political leaders. Try to build consensus in the classroom on the proposed responses. Use the three articles as examples in your discussions; that is, how would you respond to these incidences as part of a group effort.

Some observations this week:

  • It was clear that the students did not as universally read the given articles and strayed far from the task.  I decided not to steer the conversation back to the task at hand, but let it flow.  I am not sure I would do this again.  Of course, this is a common difficulty with having readings being necessary for a discussion.
  • As I wanted to get at least three rounds of small group + spokes council discussions, I tried to moderate the length of time in each: 10 minutes + 5 minutes.  I think the discussion two weeks ago was a little more interactive without light time moderation, so I am not sure I would do this again.  To me this points again to the necessity of having a longer class period.  1.5 hours would be better and possibly sufficient.

The opinions that were voiced during the spokes-council were mixed at first;  there was an opinion of protecting free speech and one that ‘damage control’ should be sought to detach the commenting individual from the institution.  There was some discussion around why these statements are still made (which points me to doing more work in presenting arguments why this continues to happen).  While a specific consensus action was not decided upon, opinions did converge on acting, noting that the statements by notable, high-profile people rises from ‘protected free speech’ to the creation of hostile environments.  This sets my mind at ease, that given time to discuss issues people will generally agree with supporting minority rights.

 

Graduate Teaching on Diversity: Week 5

I’m posting this late, but better than never:

For last Friday’s class, I adapted an exercise we use in the Faculty DPD Academy that has us reflect on the basic assumptions (and any resulting discrimination, injustice, oppression) in the production of knowledge in our given field of study or research.  Based on feedback from previous classes, I emailed out the specific discussion questions several days before class so students were able to reflect on them for more than a few minutes:

  • Who created and defined your research discipline?
    Whose perspectives were and are ignored in the development of your discipline?
  • Who funds research in your discipline?
    How does this affect the knowledge that is created?
  • How is knowledge in your discipline disseminated?
    Who has access to this knowledge and who doesn’t?
  • How is knowledge passed on in your discipline?
    Who controls this?
  • Whose interests does your discipline serve?
  • What are your advantages/disadvantages in your field?
  • Who is advantaged in your field? Who is disadvantaged?
  • Are some people systematically disadvantaged by the way knowledge in your discipline is produced and/or taught?
  • How does your discipline support and help maintain the dominant culture?
    In what ways could your discipline challenge the dominant culture?
  • What are the ethical considerations implicit in your discipline?
  • How might your discipline play a role in effecting social justice?

We used a silent discussion during class.  I posted each question (or pair of questions) on a big white piece of paper on the wall and gave each student a marker.  For 15 or 20 minutes, students went around and wrote their own thoughts;  I asked them to, for the moment, pay no attention to what other students were writing (as much as possible).  For the remaining time, students were to go around and read their peers’ thoughts, reflected on them, responded to them and added new thoughts.

I think this method of allowing students to express their thoughts worked well, but I don’t think I provided adequate readings and materials to prime the class to engage these questions in a deep way.  As I continue to discover where students are coming from, I hope to find appropriately targeted readings.  I’m finding this challenging (as I think of readings for a future offering of this course).  There are a lot of academic articles that are rigorous but also very dense.  I myself find them difficult to read, and so I’m not sure how appropriate it is for a group of students that include many English-language learners.  I am also not sure how many social science and humanities classes our students have taken (maybe I should ask) that would prepare them for this type of reading.  There are also a fair amount of pop lit to draw on (blog posts, news articles, etc), which I have been using, but I am not happy with the rigor or depth.  I’d of course welcome suggestions.

Graduate Teaching on Diversity: Week 4

For our fifth meeting of the Graduate Responsible Conduct of Research and Difference, Power & Discrimination class, I had the students read three stories that involve research misconduct or improper attribution of research.  The first describes the antisemitism and underappreciation of interdisciplinary research that lead to Lise Meitner’s not being fully recognized for her contribution to Nobel-prize winning nuclear fission research; the second describes Rosalind Franklin’s unacknowledged contributions to the DNA helix-structure discovery; the third describes a recent account of falsified research that was popularized by a This American Life story before the falsification was known.

To try to involve everyone in the discussion, I adapted a spokes council forum for the classroom.  I gave the following task:

Propose solutions to prevent or mitigate future instances of research misconduct and improper attribution. Try to build consensus in the classroom on the proposed solutions. Use the three articles as examples of research misconduct in your discussions. You should be able to argue that your proposed solutions would have prevented or mitigated these instances of research misconduct and improper attribution. You may wish to start by identifying and discussing the research misconduct and improper attribution in these stories.

And the following guidelines for a spokes council forum:

Break into small groups of at least 3 people. Each group picks a spokesperson. Alternate between small group discussions and spokes council discussions (as moderated) to build consensus around a proposed solution.

Only spokespeople talk during spokes council discussions. Spokespeople must faithfully represent and advocate the opinions and decisions of their small group, even if these conflict with their opinion. The spokesperson may change from one spokes council discussion to the next but not during a spokes council discussion (with consent of the small group). Small group members may whisper or pass notes to the spokesperson as a means for clarification.

Small group discussions should generate new ideas and reflect on the opinions of other groups as expressed during the spokes council. New arguments and counterarguments may be generated to respond to other groups in the next spokes council. Proposals should arise from small group discussions and only be communicated during the spokes council.

Consensus is not majority rule. Supporting a consensus opinion does not necessarily mean that the proposal being considered is one’s first choice. When deciding to consent to a proposed solution, you should consider the question “Is this proposal something I can live with?” rather than “Is this proposal the best proposal in my opinion?” Part of coming to consensus is to try to modify the proposal such that it suits your needs better while it remains something others can live with (or preferably agree is an improvement on the original proposal).

Spokes councils have been used for different community groups (for example) to come together on a joint issue.  I am not sure if spokes councils are used to come to consensus across a large group, but I thought this adaptation might work well to counter some challenges I’ve described in this class so far.  By allowing the students to form their own groups, I hoped that students would be among other students they felt comfortable speaking freely with.  (This seemed to work: I think I saw everyone speak at some point during small group discussions.) I hoped that all student views would be represented through the spokes council, even of those that may not typically speak up in the classroom-wide space.  We ended up spending 10 minutes in small group discussions then 10 minutes in spokes council discussions and repeating this once.  An hour-long lecture slot is never long enough!  I moderated the spokes council slightly to make sure that the spokespeople were representing the views expressed in small group discussions only; students self moderated a handful of times.  In future I may moderate the spokes council discussion for time, to allow for an extra round of small-group and spokes-council discussion.

Some interesting points and suggestions that arose include (but are not limited to):

  • A desire for positive reinforcement rather than through punishment.  The former seems more challenging to come up with; the latter easier to implement.
  • Having a bar exam and licensing equivalent for researchers, possibly through professional societies.  Concerns were raised about how you even ‘test’ for ethical behavior and how this could be implemented in a non-punitive ways.
  • All for the publication and promotion of negative results.  This was spurred by last week’s discussion, where it was discussed that the pressure to publish motivates unethical behavior.  Concerns were raised about how you would set standards for such publications.

Even though I knew that one lecture slot would not be enough time to result in consensus in one solution, I think by stating a goal and giving a method for reaching the goal, the discussion was much more in depth than in a more open-ended discussion.  A good suggestion came from a student after a class: they pointed out that it was difficult to get started having just read the articles without knowing what the purpose of the discussion was to be – I think this can be fixed by having a warm-up/reflection assignment due at the start of class to ‘prime’ the students.

From my point of view, I think this worked really well.  Or at least, met my aims.

 

Graduate Teaching on Diversity: Week 3

For the fourth meeting of our graduate diversity & ethics class, I had a bit of a break.  Three colleagues from my department generously came to give a panel on research ethics.  I asked these colleagues to emphasize the ethical considerations in the choices we make in our research questions that we pursue.  (The students are also completing online ethical training that covers plagiarism, peer review, authorship, etc.)  One panelist asked the students to read Killer Robots: The Soldiers that Never Sleep and I asked the students to read Censored Into Resignation?, Why the ‘Unhiring’ of Steven Salaita Is a Threat to Academic Freedom, This week, I resigned from my position, to try and get the students primed for considering.

The panelists made short opening statements, including considerations of funding sources (such as whether or not to seek military funding), how to decide what to research in the first place and how various factors can drive researchers toward unethical choices.  Quite a bit of time was spent discussing military funding and applications hinting at whether or not one can ensure one’s research will only be used for (the ill-defined) good.  In future years, I hope to find an appropriate article that considers the ethics of military funding and research for the students to read ahead of time.

More interesting to me is what didn’t happen.  Remember, this course is an experiment for me as I have little idea of where the students (or faculty!) are coming from.  First, there was no discussion of the corporatization of education and research, even though (I think) this has a huge impact on what we teach and study.  I also hope to find a suitable article to have this arise as part of the discussion. I think that some of the above readings touch on this, but possibly not explicitly enough.  Second, there were very few questions.  Only three or four students asked questions – all male and all with very good English.  This is an ongoing challenge, to ensure equal representation of ideas, and I have some ideas for democratizing this (for example, by asking students to each write down a question after the opening statements or based on the readings that a moderator can ask).  Third, it seems that students and faculty alike are resistant to imagining our system of research undergoing fundamental change.  I asked explicitly how one could imagine removing the influences that lead us toward unethical choices.  This might not be fair, given little warning, but I think I could design an interesting assignment around picking a part of our system that can drive unethical research and imagining a way to change the underlying system or an entirely new system that would not suffer the same problem.  All around, I have lots of ideas for next time around!

Someone pointed out after the panel, when I pointed out how few questions there were, that it is probably natural given that these students are incoming graduate students with little or no experience in research.  Of course!  And it refreshes the ongoing conversation I’ve been having with myself and colleagues across the university about the ‘right time’ to have this type of education.  If we are only given one quarter in which to engage students, one could argue that later in one’s graduate career would allow for deeper consideration of the material, with more experience.  On the other hand, at the start of your graduate program this course serves an orientation purpose – a chance to start off on the right foot.  After all (for the moment at least), I only have a 1 credit course in which to engage the students.  While this hardly does the material justice, I remind myself that it introduces our students to these concepts, it gives them the language with which to discuss these concepts, and it makes it normal to discuss ethics and oppression.

Ignorance is bliss

written on March 21, 2014 and saved for publication until tenure

I like to think that I have become more socially conscious since the onset of my adult life. I like to think that when I hear of injustice, I at the very least make the adjustments to my own life in accordance to those injustices. I limit my carbon-footprint-intensive travel to a minimum, I support unions, I no longer shop on Amazon, and so on.

It’s a learning process. I am sometimes taken aback by my own ignorance when something it first pointed out to me. Such as the day a fellow 350Corvallisite told me that half of our city’s electricity bill goes to water treatment. Half! I have always known that water use is an issue, but I had thought it more of an issue in drought-prone California. I hadn’t thought of the energy-requirements and so carbon-footprint of water use. So, I’ve started looking into gray water and rainwater collection and use.

What does this have to do with work? Well, I was sitting in yet-another-job-talk that included yet-another-use of Amazon’s mechanical turk to generate and collect data for research. The question popped into my head: “how much do these people get paid? what are the labor issues of this machine?” Given Amazon’s (abysmal) track record in labor practices, I didn’t have high hopes. Turns out they pay 50c to $5 an hour. And please don’t get me started on “its okay for someone in India to be making 50c an hour”.

I don’t know how much my colleagues are offering for mechanical turk labor — perhaps, and I hope, they offer at least minimum wage. If not already, I would hope that NSF would demand minimum labor standards for research they fund.

addendum September 1, 2014:

Ironically, it is Labor Day and I am working. A colleague just shared this “Guidelines for Academic Requestors [of Amazon Turk Labor]” with our faculty email list much to my happiness. It includes guidelines on fair payment and includes information pointing to an at-least-minimum-wage payment is required for ethical treatment of Turkers and arguments to be paying much higher than the US minimum wage standards.

Graduate Teaching on Diversity: Week 2

The third meeting of our graduate diversity & ethics class went by quickly. I broke the students into small groups of 5-6 to discuss ethical considerations of competition and mutual aid in two contexts: academic research and graduate studies – students took part in two 15min discussions – one for each context, using the following questions as a guide:

  • When is mutual aid beneficial?
  • When is mutual aid detrimental?
  • When is competition beneficial?
  • When is competition detrimental?
  • Can mutual aid result in ethical violations?
  • How can ethical violations be avoided in the practice of mutual aid?
  • Can competition result in ethical violations?
  • How can ethical violations be avoided in the practice of competition?

After the two discussions, we had some class-wide discussion which I hoped to use as a consolidation of ideas. If anything, it was very interesting to see the contrast between the small group comments and the larger conversation.  I hovered over the small-group discussions and got a sampling of the student conversations which seemed quite balanced (equal time spent on discussing mutual aid and competition) whereas the focus ended up on competition in the class-wide comments. I’m not going to take too much meaning from this, because I think I could improve my skills as a facilitator. It also seemed to me that a majority of the students actively participated in the small-group discussions, whereas only a few voices were heard during the class-wide comments (and those voiced were male dominated). I’m not sure that I could perfectly facilitate a conversation along those lines, but I think I will seek out other ways to summarize small-group discussions to the larger group that may equalize the presentation of ideas. Here the summaries were more for my own interest, since every student participated in the same discussions (and this won’t be the case in future classes). In retrospect, it might be interesting to break this conversation in two: small group discussions on competition vs. mutual aid as a student and a panel (by those more experienced in academic research than incoming graduate students are likely to be) to discuss the same idea in the context of academic research.

I had intended to talk more practically about department testing and qualifying exam requirements in the context of when “cheating” is not allowed and when collaboration is expected or encouraged, but 50 minutes goes by quickly. I would also like to have more input from my colleagues on this point, because my understanding of the expectations may not be as representative as I think. I did ask the students to listen to Computer or Human? to spur a discussion about when is your work considered your own, which would have been interesting …

The first short assignment was also due, in which I asked students for a contemporary example of discrimination resulting from research, development or technology in ECE or CS, with a paragraph each summarizing the discrimination that occurred (or is occurring) and their reaction to this discrimination. There were a number of examples of gender and age discrimination in the tech sector, with heartfelt consideration of the negative impacts. There were a few articles with an international perspective on gender discrimination, with an example of some very explicit gender discrimination, including about a law in Iran that provides more seats to men to major in engineering than woman. There were a few examples of technological development resulting in discrimination, for example in the ‘gig economy’ which can result in worse labor rights for low-income workers. Overall, it provided an interesting sampling of discrimination and student perspectives thereon. I’ll be asking for a similar submission at the end of quarter, but will reword the ask to push the students a little further.

Graduate Teaching on Diversity: Week 1

I had the second class for our pilot graduate diversity & ethics class; I discussed the first class last week. I asked the students to read Leaning In: A Student’s Guide to Engaging Constructively with Social Justice Content and to review the meaning of 16 words that we can use to talk about diversity, discrimination and oppression.  Since these words are new to most students (I think) and even just their meanings can be difficult to internalize, we did an exercise that I picked up in our faculty difference, power and discrimination seminar.  Students were each assigned one word and wrote down a definition.  Students then paired up and had 3 minutes to describe their definitions to each other. Now each student had 2 definitions.  They paired up again and had 5 minutes to describe their 4 definitions to each other.  They paired up a final time describe their 8 (or so) definitions.  We then had an open discussion about some of the more challenging definitions (if I remember: institutional vs. structural discrimination, internalized oppression, hegemony).

identity-mapI then had the students draw an identity map and to explain I showed them mine (right), and of course admitted that this is by no means complete (two glaring omissions are religion and body-type).

My goal with all this is to have students realize the privilege they are granted based on their identity in addition to simply thinking about one’s own identity.

Between talking about the diversity terminology and talking about identity, discrimination and privilege, there was a fair amount of discussion and some resistance to admitting access to privilege as well as denial of discrimination based on other identities.  We talked a little bit about student resistance at the faculty DPD seminar, but I can’t say I was ready.

I remember the first classes I’ve taught.  I remember the times that I got stuck in a lecture or made a mistake or was unable to explain myself.  And Friday’s class wasn’t dissimilar.  However, I’m a trained theoretical computer scientist, so when teaching technical material I can draw on years of experience and know that I can probably do better next time; with this graduate diversity teaching, I don’t have nearly the training.  There is also a lot more emotion tied up with ideas of discrimination and privilege.  It is really difficult to see yourself as perpetuating systems of oppression.  It’s important to not feel blame, but I think it is important to recognize that we are all a part of these systems of oppression and we are all responsible for fighting against it.  I’m not sure I did a good job of explaining that on Friday,

So, I left class with a lot on my mind and woke up the next day with it still weighing heavily.  I only have 10 weeks/hours to discuss these ideas explicitly with these students and I can’t help but blame myself for not doing the best job possible.  Discussions this weekend with my partner, a colleague in Philosophy and (very welcome emails) with a student in the class have helped to digest this.  I have a lot to learn and I am ready to accept that the first run through of this class is not going to be perfect.

I just hope I have another opportunity to teach this in the future.

Raises, Salaries and Greediness

written on July 3, 2014 and saved for publication until tenure

Someone told me once that Knuth turned down raises at Stanford and that this led to problems because they had rules about salary inversions — that it prevented Stanford from offering starting salaries higher than Knuth’s which was, presumably, modest. I might be misremembering, and I couldn’t confirm this with the all-trustworthy Internet, so it might not be true

However, in my searching, I did read some Knuth quotes and stories that made me feel very good about the world. That there are people out there who aren’t greedy, don’t want to try to monetize everything, and just want to do good. So, here they are:

Knuth decides to not be a compiler writer for the rest of his life and decides to focus on what is important in life:

“Then a startup company came to me and said, ‘Don, write compilers for us and we will take care of finding computers to debug them. Name your price.’ I said, ‘Oh, okay, $100,000,’ assuming that this was outrageous. The guy didn’t blink. He agreed. I didn’t blink either. I said, ‘I’m not going to do it. I just thought that was an impossible number.’ At that point I made the decision in my life that I wasn’t going to optimize my income.”

“The important thing, once you have enough to eat and a nice house, is what you can do for others, what you can contribute to the enterprise as a whole.”
(From Jack Woehr. An interview with Donald Knuth. Dr. Dobb’s Journal, pages 16-22 (April 1996))

“I decry the current tendency to seek patents on algorithms. There are better ways to earn a living than to prevent other people from making use of one’s contributions to computer science.”
(Donald E. Knuth, TAoCP vol 3.)

Maybe Knuth is an anarchist too.