English (United Kingdom)Russian (CIS)
Добро пожаловать на сайт Международной Гильдии Лидеров Перемен и её Управляющего Центра. МЫ НЕ КОНСАЛТИНГОВАЯ ФИРМА. Мы - союз SbA-практиков управления изменениями, центр компетенций, аккумулятор знаний. Будущее за Лидерами Перемен. Если вы - практик, давайте сотрудничать. Если вы генеральный директор - начните проект улучшений с его обсуждения с нами.


Сначала мышление

Блог для практиков управления изменениями

Анатолий Вассерман: Как

отличить правду от лжи

Как отличить правду от лжи

Черниговская бизнесменам

Татьяна Черниговская: интервью для менеджеров

ТРИЗ Альтшуллера

ТРИЗ Альтшуллера


Чего вам не хватает на этом сайте?

3 Lessons I Learned from John H. Holland
Статьи - Познание (меметика и другие)
Автор: David E. Goldberg   
20.09.2017 12:00

Джон Генри Холланд - John Henry HollandДжон Генри Холланд (англ. John Henry Holland; 02.02.1929 — 09.08.2015) — профессор психологии, электротехники и информатики в Мичиганском университете, Анн-Арбор (США). Один из первых учёных, начавших изучать сложные системы и нелинейную науку, известен как отец генетических алгоритмов. Статью о нём ученика — Девида Голдберга — предлагаем вниманию коллег, читающих на английском.

David E. Goldberg, 20.08.2015

Complexity science pioneerJohn H. Holland passed away on Sunday, 9 August 2015 (here). Those of us who got to work with him were blessed, and, like so many of his other students, my life was irrevocably altered by working with him. To remember John, I’d simply like to tell the story of how I met him, and three things I learned from him that had roots in that first encounter.

Meeting John

It was the fall of 1980, and I had returned to Michigan to get a PhD in Civil Engineering. Prior to returning to school, I worked writing and installing real-time pipeline simulation software in oil and gas pipeline systems, and I had the realization that long-distance gas pipeline operators “drive” pipelines like you or I drive a car, so I returned to school with dreamy visions of doing artificial intelligence; I signed up for the standard CS course on Artificial Intelligence, went to class, and found a sign on the door saying that the course was cancelled.

In seeing the sign, all I could think was that my dreams of doing AI and pipelines were being crushed by the cruel fates of class scheduling. I looked through the course catalog searching for a class replacement and I came across a course called Introduction to Adaptive Systems, CCS 524, taught by someone named J. Holland. It wasn’t exactly what I was looking for, but it was the best game in town, so I signed up and went to class.

I arrived in the classroom, and standing at the front was an energetic and youngish looking prof. He said the class didn’t have any exams, it had one term paper/project on any subject, and two course books. One book was something called Adaptation in Natural and Artificial Systems, a text written by the instructor himself in 1975. The second book was a collection of papers published in 1963, Computers & Thought. I couldn’t help wondering at the time whether there hadn’t been anything more interesting published in AI than a 17-year old collection of papers, but I reserved further judgment until I could get the books.

The youngish prof put on his glasses and started talking about what sounded to me like a randomly selected and unrelated series of subjects: genetics, economics, automata theory, schemata, Samuel’s checker player, some strange construction he called a classifier system. I didn't understand what this had to do with “real AI” or adaptive systems, but the prof seemed earnest enough, and there was a deep confidence about him. Nonetheless, I couldn't fathom how any of this might lead to a dissertation in Civil Engineering.

This first experience was perplexing enough, and I thought matters couldn’t get any worse, but I was wrong! I went to the bookstore, got the two texts, and took them home. The collection of papers was dreadfully old and outdated, it had no working code, and it even even smelled bad—musty and old—like a flooded basement after the water recedes.

And the author’s book. Yikes!! It was filled with equations that didn’t relate to anything I had experienced in my equation-filled engineering education, and it moved from what at the time seemed like disparate topic to topic with what appeared to be reckless abandon.

What had I gotten myself into?

But, I stayed with the course, wrote a dissertation applying genetic algorithms and classifier systems to gas pipeline control, and had the privilege of learning from and working with one of the early masters of modern complex systems.

3 Lessons I Learned from John

I love telling this story because of the serendipity and intellectual tension of the first encounter. As I reflect back, one of the puzzlers is why I stuck with the course; it met none of my prior expectations, and yet I stayed. Of course, I am glad I did. Doing so changed the course of my life, but what was it that kept me coming back for more?

I think I stayed, in part, because I had an intuitive sense of some of the great lessons in store by working with John. Here, I'd like to summarize three of those lessons.

Lesson #1: Tell Great, Coherent Stories

One reason I stayed with the course is because John was a great teller of science stories. Sometimes his equations were hard to follow, but I always got his story, and it was the coherence of his narrative that convinced me that his stuff could be made to work.

Since that time, my work as an engineer, scientist, and now as a leadership coach has led me to think of stories as the central way in which human beings approach complexity. Stories contain, time and spatial relationships, causality relationships, intentionality relationships, clues regarding ontological modes & epistemic certainties and uncertainties, all in a compact representation with the possibility of pointers to other stories as well as visual and mathematical representations. I couldn’t have had a better introduction to the importance of great narrative in a scientific context than listening to John.

Lesson #2: Have the Courage to Jump Long

One of the puzzles for me in meeting John was getting used to his bringing insight from so many different fields. He did it so effortlessly and naturally, yet for a young engineer it was bewildering to see someone jump outside of his own discipline with such confidence, always bringing back so many useful things.

Since that first day in class in 1980, university life has become marginally more interdisciplinary, but students of John smile knowingly at each other when others talk about interdisciplinary work. John taught a kind of extreme courage to go wherever you need to go in the interest of doing good work. Rarely do you see it done with such grace and aplomb as by John, but many of us touched by his example are better able to jump longer and further afield than we otherwise would have without his leadership and example.

Lesson #3: Trust Yourself & Others

I think another thing that kept me in class during those first days and weeks when I was still full of doubt and skepticism was John’s demeanor. He was completely at ease and quietly confident about the things he was saying, and it wasn’t arrogance. It was what coaches might call “leadership presence.” He was connected with us in class. You trusted what he was saying even if you didn’t fully understand it, and he was curious about our questions and reservations. It was attractive in a way that was and is hard to describe.

This attitude carried over to the way he “managed” us as his PhD students. Basically, he trusted us to figure things out. He would listen carefully to our results, he would ask a few questions, but he rarely was directive about what should be done next. At the time, I think I wished for a little more advice. In hindsight, I am grateful he did things as he did.

The Last Time I Saw John

I last saw and spoke with John in early 2013 in Singapore at an event at Nanyang Technological University (NTU) commemorating his 85th birthday. I was in Singapore on other business and traveled across town to listen to John lecture on portions of his 2012 book, Signals & Boundaries. The vibrancy of his storytelling, the courage of his long jumping, and his presence & trust came through as before; and this particular talk sent me off on a weeklong reflection regarding the concept of lever points in ways that felt familiarly like reflections so many years earlier.

During a break, I pulled John aside, and I was moved to tell him how much he had meant to my life, and how grateful I was for having been his student. He smiled and moved on to a pressing appointment. Thinking back about that last encounter, he didn’t seem that much different than the youngish looking prof I met 35 years ago. And maybe in that final observation is another lesson for us all.


David E. Goldberg is best known for his work on genetic algorithms and his first book Genetic Algorithms in Search, Optimization, and Machine Learning (1989). Today, he is a trained leadership coach (Georgetown University) and president of ThreeJoy Associates, Inc., a coaching, training, and change leadership consulting firm in Douglas, Michigan. He resigned his tenure and a distinguished professorship in 2010 to work full time for the transformation of higher education. He can be reached at Этот e-mail адрес защищен от спам-ботов, для его просмотра у Вас должен быть включен Javascript

This is an abbreviated and modified version of a piece that appeared in Aha... that Is Interesting, a book of short papers honoring John on the occasion of his 85th birthday.

О Джоне Генри Холланде в Википедии