It's different when you're big
© 8 September 2021 Luther Tychonievich
Licensed under Creative Commons: CC BY-NC-SA 3.0
other posts

Reflections on why leading 100 is not like leading 5 groups of 20.


Four components of my recent life have all caused me to reflect on the same principle, a principle I didn’t understand in my youth.

In seminary I’m teaching about the revelations received during the time when the church was scaling up from a few hundred to a few thousand members and was starting to add more structure: multiple groups of leaders with assigned responsibilities and reporting lines, rules for handling various kinds of problems that were becoming common, and so on.

At work, I am engaged in several different policy matters, helping document structures and practices that have previously been undocumented and consulting on how written policies could be made clearer, more precise, cover extra cases, or leave some aspects unspecified.

As part of my external service for the university, I am working with various parties associated with to better support researchers and departments in creating and executing Broadening Participation in Computing plans. One part of this conversation is how to change some text that is understood by our dozen or so BPC consultants in such a way that it will be understood the same way by a much broader audience.

Next semester I will be teaching Computer Systems and Organization 1. I previously taught this course twice as the sole instructor with fewer than 80 students each time; next semester I will be one of 2–3 instructors teaching several hundred students, and I’ve begun reflecting on how that structure will change how the course works.

Each of these has a component of “‍things are different when there are more people involved.‍” And that thought always brings to mind one of my favorite quotes from Cliff Stoll’s fascinating book, The Cuckoo’s Egg. In the post-epilogue, which I understand to have been written in 1990, we find this exchange between the author and some person named Dave who had attacked the computers of the Smithsonian Observatory, where the author was then working. We pick up with Dave’s statement:

“‍I broke in to show you that your security isn’t very good.‍”

“‍But I don’t want to secure my computer,‍” I replied. “‍I trust other astronomers.‍”

Dave had other reasons for breaking in, too. “‍You think that hackers are bad. This proves otherwise.‍”

“‍Huh? You broke into my computer to show that hackers are good?‍”

“‍Yeah,‍” Dave replied. “‍We’re helping you out by finding your security faults.‍”

If you tell some kids they’re behaving like children, they’ll get even with you by acting like children.

I have used that passage to explain multiple principles in the past. Today I am thinking about it as a portrait of the changing attitude toward computers that Cliff Stoll lived through. At the beginning of the book, he describes a world where people with networked computers were few and rules weren’t much needed. Most of the book describes how he accidentally helped expose that some computers had assets that needed more protection. And then in the post-epilogue we see him wishing he still lived in the simpler smaller world even as he starts to see evidence that even a few vandals and immature people can break everything and security policies are becoming necessary for all.

The CS department I work in has roughly doubled in faculty and quintupled in student body since I joined it. Back then many situations were rare enough that the most effective policy was “‍handle it case-by-case‍”. This was both a more effective use of faculty time than deliberating policy for rare eventualities, an could be relatively fair as well because in each case the few related cases could easily recalled and considered to ensure a comparable standard was being applied. As we’ve grown, more and more situations have gone from unique to part of a pattern to being fitting a semi-formal rule of thumb to being governed by an official policy. We never set out to be a policy-driven department; we were thrust into it by necessity caused by increased size.

The changes needed at scale are not always well received.

When the church had 100 members, it had two officers: the first and second elder; and any member could appeal to the first elder at any time on any topic. The second elder was less than delighted to lose that title and become instead one of twelve high councilors and one of more than fifty church officers. Some members were less than happy to learn that added structure meant their appeals were now heard by a council of elders and not the first elder himself. The first elder, on the other hand, was relieved to be relieved of some duties and find support in new structure.

When a new policy is being discussed in our department, there are certain faculty I can rely on to always make a comment like “‍Why do we need this policy? Let’s just let each faculty member do their own thing like we always have.‍” Or, to paraphrase Stoll, “‍I don’t want to secure my policies. I trust my colleagues.‍” While there is benefit in the encouragement to simplify and be less prescriptive, in the end the dream of small-scale freedoms at a larger scale is just that: a dream. A daydream for most, a nightmare for those with the burden of leadership.

When I teach a course of 50, I don’t need a late policy or a makeup policy or an extension policy or a prerequisite enforcement policy: I can handle all of those on a case-by-case basis. If I make a too-eager deadline or many students have a deadline conflict with another class, I can find that out and adjust things on the fly, minimizing the ripple effects of then change and helping stabilize those rocked by its waves. When I teach a course of 500, individual case-by-case handling becomes more time-consuming than I can manage and synchronizing changes across sections and instructors means even simple changes require thought and conversation before implementation.

More people demands more structure. The nature of that structure can be changed, but the structure itself is necessary. Sometimes it’s happy structure, like when we added a regrade request process to a large course and found the result was that grading errors were being corrected for all the students instead of just the most vocal few. Sometimes it’s less pleasant, like Stoll’s realization that “‍If you tell some kids they’re behaving like children, they’ll get even with you by acting like children‍” and you have to add security to keep vandals from interrupting science. But happy or not, structure follows size.

Looking for comments…

Loading user comment form…