Editor's Note:

Question: What do Gregory Bateson, Niels Bohr and Shannon's Law have in common?

Answer: They all figure prominently in Lewis Perelman's thinking on technology, information, knowledge, and learning. We caught up with Lewis in his office in Virginia, where he is busily at work on a newsletter on knowledge.

Bureaucrats be forewarned: What follows may be hazardous to your thinking.

• Lewis Perelman's Biography

JOB: You stated in your book, "School's Out," that the term "information age" is outdated, and that what we are witnessing is only a transitional period, a prelude to a new historical epoch that should be more accurately labeled the "knowledge age." Why is that?

LJP: There is an important distinction between information and knowledge that is only beginning to seep into people's consciousness. This is part of the reason that we created our newsletter, Knowledge Inc.. People in business have come to realize that knowledge is important to modern management, without knowing yet how or why or even exactly what corporate "knowledge" is.

JOB: And how have you found answers to those questions?

LJP: I've been professionally concerned with studying and working on the ecology of knowledge, learning, and their role in social processes for nearly three decades. But you could say it has been a lifelong interest. Along the line I found some of the most valuable insights in the work of Gregory Bateson.

JOB: Who is Gregory Bateson?

LJP: Bateson was formally an anthropologist who did profound work in several fields, from computer science to psychotherapy. His book "Steps To An Ecology Of Mind" to me was and is a breakthrough. Ironically, I discovered it in the health food section of a supermarket at a moment when I was acutely stymied in my doctoral dissertation. It was an experience almost like something from the pen of Jorge Luis Borges: A magic book that put into words exactly the vision I had been groping to express.

JOB: So you bought the book in more ways than one.

LJP: Yes. It's a shame the book is out of print. Bateson was one of the great minds of the 20th century. He and his wife Margaret Mead did pioneering work in anthropology. They were the first anthropologists to use film as a scientific resource. Bateson had an uncanny gift of observation.

JOB: What did you find in his "Ecology of Mind" that fits in the current interest in knowledge-age business?

LJP: One of the major themes in Bateson's work was the importance of the Law of Logical Types, which comes out of Whitehead and Russell's "Principia Mathematica."

JOB: What is that law?

LJP: It says that it is a fundamental logical error to mix a class with its members. In simple terms, it is a mistake to eat the menu instead of the dinner. Much of the work Bateson did flowed out of recognizing how commonly individuals and societies got into trouble by violating this law, often with catastrophic consequences.

JOB: I have often thought that 99.9% of the world's population is ignorant of basic economics.

LJP: It is bad enough how little people know about economics. Unfortunately, most know almost nothing about knowledge. That's one barrier to satisfying Peter Drucker's call for a knowledge-based economics. There is nothing in the schools that requires anybody to have the slightest understanding of how knowledge, thinking, and learning work.

JOB: How does this jibe with the basic assumptions of economics?

LJP: Neoclassical models of economics assume rational behavior. The flaws in that approach are being exposed by exciting new work in economics that gets at actual human behavior. We now know that markets work very differently when you change the assumptions and look at how people really behave as opposed to how some Platonic automatons behave.

JOB: I think we took a nasty detour on the path to knowledge years ago by specializing the sciences. If you accept this, then you have to be encouraged by the work coming out of places such as The Bionomics Institute and the Sante Fe Institute--work that involves linking or networking various disciplines in dynamic models that advance our understanding of the real world.

LJP: Artificial life and other new simulation techniques offer a rich opportunity, but it is an opportunity that easily can be mishandled. In fact, it's an extension of Murphy's Law that these methods mostly will be misused or misunderstood. I was heavily into modeling for a big swath of my early career. I started as an applied mathematician and physicist. For a while, I was a disciple of Jay Forrester and system dynamics. At the time, I thought it was very useful. However, in a 1980 paper I published called "Time in System Dynamics," I contrasted policy models hatched in the Cambridge academic cocoon with my experience in the real world where decisions actually get made. Other critics had argued that Forrester's kind of deterministic models were unrealistic because they ran the same way backwards as they did forwards. Of course, the real world doesn't work that way, because of thermodynamics, chaos, Heisenberg, a lot of things. The result of this is that, on the one hand, you get a very idealistic view of how social systems work, and on the other hand, the criticism made early on was that these models could not account for the creativity of human behavior--of people thinking of new ways to deal with evolving situations. Genetic algorithms and other new methods may handle those things better. But the point I emphasized then and in a later paper I did at IIASA [the International Institute for Applied Systems Analysis] is that the human politics of policymaking overwhelm any methodology.

JOB: So you don't think all the new modeling power that has evolved out of advances in information technology will allow us to learn more?

LJP: I think there are some things we are going to be able to learn better, but there is going to be a lot of garbage. Many people will not use these nifty new tools appropriately. They will over-use them and stretch them to do things they were not designed to do. Furthermore, underlying all of this, the problem with any kind of modeling is that ultimately a model is only a model, and not reality. And this gets us back to Bateson's logical types and the menu/dinner fallacy called "reification." This is very closely tied to another very little understood principle of the modern world, which is Niels Bohr's complementarity principle. I don't think one in 10,000 people know what that is either.

JOB: I am one of those 9,999. Explain.

LJP: You probably learned it in high school, but like most people, forgot it after three months. It is actually one of the most important ideas of our time. Bohr said that there is a complementary relationship between our verbal models of reality and reality itself. Reality doesn't have to obey our vocabulary. Just because we don't have the right words doesn't change how the world works. The problem is in our vocabulary, not in the universe. The fact that we don't know how to express our understanding of things like quanta, quantum mechanics, and relativity--that this is so non-intuitive from the viewpoint of our average daily experience--doesn't mean the universe is wrong. It just means our language is imperfect. This is an another form of the idea that Bateson applied to many other things, from psychotherapy to alcohol treatment to human ecology.

JOB: Let's get back to the distinction between information and knowledge and learning.

LJP: Bateson dealt with this in his book and he talked about different levels of learning, from 0 to 3 based on the law of logical types, where each is a classification of the previous one. But people don't understand this. They keep mixing this up and it causes all kinds of problems. Learning level 0 is basic programming, robotic behavior if you will, where you do something over and over again with no change. Learning level 1 of learning is a system which is able to alter its programming based on feedback. And then beyond that, learning level 2 is altering that, changing the process of learning in response to the context. That level jibes with what we now know about multiple intelligences and styles of learning.

JOB: There are some talented people out there today, such as Federico Faggin, who are trying to design chips that learn. Any thoughts on this topic?

LJP: All it probably takes is a sufficient degree of complexity. If you had a complex chip with enough logic gates, you could make fairly sophisticated neural systems. If you get the densities high enough, you begin to get to the order of complexity of a brain.

JOB: If we only knew what Mother Nature already knows!

LJP: During the next century, artificially intelligent organisms will be a fact of life. People are going to have to cope with this.

JOB: How do you define knowledge?

LJP: Good question. Bateson provided about as simple and powerful a definition of these things as anybody I have read anywhere. Bateson made the distinction between a message and information, and then between information and knowledge or learning. According to Bateson, a message is a difference: marks on paper, blobs of magnetism on a disk, electronic switches -- anything that can represent a zero/one difference. Information is a message that makes a difference. He noted that a message only becomes information when it is communicated -- which means the message has to be perceived by someone, it has to result in some kind of tangible change. That is Shannon's Law of Communication. By definition, unknown data is utterly useless.

JOB: How does information become knowledge?

LJP: Let's go back to the distinction between the information age and knowledge age, particularly from a business point of view. For the last 40 years, we have been living in the information age. We have been pumping more messages, in Bateson's terminology. But that is not knowledge. Data has to be perceived to become information. Then, a step up in logical type, knowledge is a difference that makes a difference in how information is perceived. That process of changing how information is classified or used is also called learning. So the difference between information and knowledge is the difference between awareness and meaning, between being informed about something and knowing how to connect it to something else. Simply, it's the difference between know-what and know-how.

JOB: Drucker often harps on this point. He argues that the CEO needs knowledge, not just information.

LJP: I agree with what Drucker said in his book, "The Post-Capitalist Society," that the next big wave in economics is coming up with the theory of a knowledge-based economy. It's a big problem that is mostly unsolved. The handicap of Drucker's work is vocabulary. He is talking about a new kind of schooling or reinventing education but he still uses the vocabulary of schooling and education. This is one of the basic problems with education today. Education is stuck at the level of having people memorize material and regurgitate it. You are considered a well-educated person if you are a well-informed person.

JOB: My mentor at graduate school used to tell me that the first step to knowledge is knowing what you don't know.

LJP: Well, that is true, too. You see, that is the problem. The truth is we have no way of literally knowing what anybody knows. It gets back to the distinction between data and information. Bateson said information is a process, it is a loop, it is not a location. All we can observe is how somebody performs something. We give them a test. This is not the same as looking into a brain and seeing what is stored there. We can see brain activity with probes like PET scans, but we can't yet see what anybody knows. We may someday be able to, but we certainly can't today.

JOB: I wonder what kind of valuation that kind of knowledge would fetch on the street?

LJP: It reminds me of one of Woody Allen's early jokes. He said he took a speed-reading course and afterwards he was able to read "War and Peace" in forty-five minutes. "It was about Russia," he said. That's funny because there is an important truth in it. I took one of those courses that promised I could increase my reading speed by five times with what they called "greater comprehension." I found that meant "remembering" what you had read. But I don't read things just to remember them. My clients want to know what I think about what Michael Rothschild or George Gilder or whoever wrote. I'm still a slow reader, not because I can't scan words quickly but because I need to spend time thinking about what the writer is saying. The speed-reading business rests on another one of those menu/dinner errors: confusing "recollection" with "comprehension."

JOB: Exactly. Makes me think about these aptitude tests that educators constantly use as measures of knowledge to screen individuals. They are probably better measures of recall. But that isn't knowledge, is it?

LJP: Again, one version of Bateson's learning level 2 is learning how to learn. That is a very popular idea today. He further added a profound insight: that learning level 2 is probably what we mean by personality. Character and personality are ways of organizing the processes of learning how to learn. Bateson also speculated about what learning level 3 would be: learning how to learn how to learn. He suggested that might appear as a basic personality switch, even the spiritual "born again" kind of experience. You have to think about that a lot to see what he is talking about there.

JOB: Hmmm.... Learning how to learn how to learn.

LJP: Right. Like Boris Yeltsin suddenly realizing Communism sucks. Like Bill Gates realizing that the Internet is important after all. It is more than just learning how to learn. It is something that changes the whole structure of perception--the paradigm switch. There is no mathematical limit as to how many levels there can be. There is no intrinsic boundary. There must be another level beyond that. But then you are probably getting close to something like God.

JOB: What are you working on these days?

LJP: I am working primarily on our newsletter, Knowledge Inc., which we started in April of this year.

JOB: What kinds of issues are you thinking about?

LJP: A lot of different issues, but many are related to the Law of Logical Types. You know, the big tragedy of modern civilization is that when Moses came down from Mt. Sinai, he thought he was delivering a prescription, and the children of Israel thought they were receiving a menu. This is the frustration of many business consultants today, from Tom Peters to Michael Hammer, myself included. It is the dilemma that the Soviet world worked its way through. Reform doesn't work. One of the dilemmas of reform is the more you succeed locally the more you fail globally.

JOB: That is powerful.

LJP: Or as Kenneth Boulding put it more succinctly, nothing fails like success. And that is the delusion of all reformers. As I reported in my book, most of the many research studies of reform efforts -- across business, government, military, religious and other institutions -- show they are almost always failures. And the reason is readily understandable in a bionomics framework: Every living organism has some kind of an immune system which is programmed to attack anything alien. And of course, the more successful the invader, the more vigorous the counterattack will be. Successful pathogens usually are the ones that exploit the immune system's blind spots or Achilles heel. But reformers generally prefer the frontal attack, which strategists from Sun Tzu to Liddell Hart have shown usually is self-defeating. Technological innovation often founders on the same rocks. History is littered with inventors who had superior technology but who were deservedly forgotten because they did not know how to commercialize it.

JOB: No doubt about that. Apple comes to mind, as does the Betamax.

LJP: I know Apple freaks who just despise Bill Gates because he has sold junk technology. What does he care? He is worth about $20 billion! He understands what Alfred Sloan said, that General Motors is not in the business of making cars, GM is in the business of making money.

JOB: I remember a great quote from your book by Theodore Levitt who said that people don't buy quarter-inch drills because they want quarter-inch drills--they buy quarter-inch drills because they want quarter-inch holes.

LJP: It all goes back to my guru Bateson and the Law of Logical Types. If you don't understand this concept -- and most supposedly "educated" people have never heard of it -- you may spend your whole career in a state of perpetual confusion.

JOB: Don't eat the menu, right?

LJP: Look, if you want to eat the menu, you can. That is your choice. Most of the dilemmas and problems with our economy really flow from this problem. Another name for this, by the way, is the fallacy of misplaced concreteness.

JOB: Can you relate this to bionomics?

LJP: Actually, one of its many manifestations is the basic law of genetic evolution, which prohibits the inheritance of acquired characteristics. This is just another example of the Law of Logical Types which Darwin understood and Lamarck did not. DNA is information. Experience is also information, right?

JOB: Right.

LJP: Why can't one be combined with the other? Because one is a classification of the other. Memory data are what an organism does. DNA data are what an organism is.

JOB: I always liked the quote attributed to Aldous Huxley who said that experience is not what happens to you, but what you do with what happens to you.

LJP: Yes. This is still another example of Logical Types. Think about the problem of unemployment. Our employment system still runs basically on the resume and academic credentials which are accumulated information about what happened to you. The resume doesn't say anything about what you can do. It is a different logical category. Think about the cost of that error. As you know, I have written about this extensively. We could totally transform our economy if we could just get employers to consistently employ people based on what they actually can do.

JOB: Not on what a resume says they can do?

LJP: Hire performance. That is what you want. That is the dinner. Yet what many employers buy is the damn menu!

JOB: Don't those credentials say something about your future performance?

LJP: Not as much as what most employers think. In fact, in the knowledge age enterprise academic achievement is often a negative correlate of work performance. I mean, the menu says something about the dinner, too. But anybody who has ever had any experience in restaurants knows that no matter how fancy the menu is and how high the prices are, there may still be roaches in the kitchen. In fact, the Washington Post did an expose several years ago which revealed that the poshest restaurants in DC were often at the top of the list of health code violations.

JOB: We will remember that when we dine in DC the next time! Any other pearls of wisdom before we wrap this up?

LJP: I just want to reiterate that the difference between information and knowledge is the difference between humans performing or not performing in the context of a specific situation. One of my favorite Sherlock Holmes stories provides a good illustration. Holmes and Watson are in the parlor of a house where a murder has taken place. They are both looking around examining the scene of the crime, and Holmes says to Watson, "I see it all now, I know who did it." And Watson -- who, after all, is a trained physician and no dummy at diagnosis -- says with his usual astonishment, "My dear Holmes, I've examined this same room with you and I see nothing at all!"

And Holmes' repost is, "No Watson, you 'see' everything, but you 'observe' nothing." In other words, knowledge is the ability to connect information, to discern what Bateson called "the pattern that connects."

That was Holmes' special skill: to find meaning in a mass of seemingly trivial data. Watson was intelligent, but Holmes was smart.

JOB: Thanks for your time, Lewis. Good luck with the newsletter.

Lewis Perelman is principal of the Perelman Group and executive editor of the management newsletter Knowledge Inc.

Related Links:

• Perelman Group
• Knowledge Inc.