Creator of Fuzzy Logic
Lotfi Zadeh in his office at Berkeley. September 1994.
Lotfi Zadeh was born in Baku, Soviet Azerbaijan, in 1921; today, his scientific concept, Fuzzy Logic, has strongly impacted computer technology, tomorrow it may be shaping the way we perceive the world.
Back in 1965 when you published your initial paper on Fuzzy Logic, how did you think it would be accepted?
Well, I knew it was going to be important. That much I knew. In fact, I had thought about sealing it in a dated envelope with my predictions and then opening it 20-30 years later to see if my intuitions were right. I realized this paper marked a new direction. I used to think about it this way-that one day Fuzzy Logic would turn out to be one of the most important things to come out of our Electrical Engineering Computer Systems Division at Berkeley. I never dreamed it would become a worldwide phenomenon. My expectations were much more modest.
Lotfi Zadeh (2nd from right) returned to visit Azerbaijan, his country of birth, in 1967. Here he is seen with officials at the Azerbaijan Oil Institute. In 1993, the Azerbaijan Academy of Oil honored Lotfi with an Honorary Professorship.
How did you think Fuzzy Logic would be used at first?
In many, many fields. I expected people in the social sciences-economics, psychology, philosophy, linguistics, politics, sociology, religion and numerous other areas to pick up on it. It's been somewhat of a mystery to me why even to this day, so few social scientists have discovered how useful it could be. Instead, Fuzzy Logic was first embraced by engineers and used in industrial process controls and in "smart" consumer products such as hand-held camcorders that cancel out jittering and microwaves that cook your food perfectly at the touch of a single button. I didn't expect it to play out this way back in 1965.
How did you come up with the term, "Fuzzy Logic"?
I coined the word "fuzzy" because I felt it most accurately described what was going on in the theory. I could have chosen another term that would have been more "respectable" with less pejorative connotations. I had thought about "soft", but that really didn't describe accurately what I had in mind. Nor did "unsharp", "blurred", or "elastic". In the end, I couldn't think of anything more accurate so I settled on "fuzzy".
Would you say that Fuzzy Logic turns Aristotelian or Classical Logic on its head?
(Laughs). Back in Aristotle's day, people tried to be as precise as possible. That's the Aristotelian tradition, the Cartesian tradition. Looking at things as being entirely black or white stems from such a tradition. But take the example of good and bad. What we're beginning to understand now is that sometimes things that we perceive as bad really turn out to be good, or perhaps, not as bad as we originally thought. Things can serve a purpose. People back in Aristotle's time and even later thought that by perceiving things in black and white (in absolute terms) that they gained alot. And they did. But they lost a great deal in the process. Fuzzy Logic represents a swing in the opposite direction but I would like to stress that there is much more to Fuzzy Logic than multi-valuedness of truth.
Classical logic has erred in devoting so little attention to approximate reasoning and focusing to such a high degree on exact reasoning. So when you take a course in logic, you learn all kinds of things which are of very little use in everyday life. We encounter approximate reasoning all the time. For example, "Where can I park my car?" Where should I have lunch? Should I place this call "person-to person" or "station to station"? Should I buy this house? How do I get from this side of town to the other when I'm in a hurry? Classical logic, operation research, decision analysis-many other disciplines have nothing to say about this topic.
How could they ignore it?
When the only tool you have is a hammer, everything begins to look like a nail. Classical logic simply doesn't provide the means to solve the problems. They concerned themselves with models of precise knowledge. But such models are so far removed from the real world that they don't do you any good.
There's a field called "Game Theory." Hundreds of books and thousands and thousands of articles have been written about it since the late 40s. Now you might think that by reading those books and articles and by taking courses, you would be in a better position to play games of various kinds. But the truth is, it doesn't do you a bit of good. Not one iota. Game Theory is concerned with models that are not tied to real games and only tenuously tied to real world conflicts.
Why are you so "down to earth" in your own approach to solving problems?
I'm that kind of person. I've been conscious for a very long time that the real world is very complex. There's always a tendency to oversimplify. That was the case in physics. In my youth, we had very simple models. About electrons in orbit. About protons. Models of the solar system. Of course, today we realize that these were ridiculous. The physics of these phenomena are infinitely more complex.
Today we do the same with neuro-systems. We think that there is a close similarity between those systems and what goes on in our heads. But that's not the case. There's a big gap. There are big gaps in many fields.
Why is Fuzzy Logic described as "cheaper" and "easier" than traditional methods of computing?
Fuzzy Logic is "coarse". The important thing about Fuzzy Logic is that it's done in a "coarse" way-not refined. Anything that is coarse-grained is simpler and cheaper. If, for example, you want to park a car and somebody said, you had to do it within ±1/10 of an inch of some particular point, you'd be in trouble. The reason why people can do things like park a car is that they don't have to be very specific to succeed.
Many people don't realize that this is one of the very important features of Fuzzy Logic. You use what I call "granulation" which means you lump things together. It makes things easier, cheaper and faster. If you had a bunch of screws and nuts and you picked up one screw at a time and took it some place and then came back for another, it would take a long time. Dump them together in a bag and it's faster. That's what you do with Fuzzy Logic.
Everybody is talking about the Japanese being so advanced in Fuzzy Logic. Is that true?
It's true. The Japanese are very consumer oriented. They are incorporating Fuzzy Logic into many of their products, especially appliances and electronic equipment. But I just returned from Germany a few days ago (October 21st) and was very impressed by what is happening there. There are many, many people working on Fuzzy Logic both in industry and the universities. There are many good papers and books. I would say that next to Japan, Germany is the country with the highest level of activity.
And where would you place the United States?
Third. Russia used to be high up there and there used to be quite a few scientists interested in Fuzzy Logic and a number in Azerbaijan. But with the breakup of the Soviet Union, there's no money for these things now.
China is very active, too, isn't it?
The Chinese Government used to be very supportive of Fuzzy Logic. Scholars there have written quite a few papers related to traditional Chinese medicine, mathematics and engineering. You often hear a figure of 10,000 Chinese scientists involved with Fuzzy Logic. That figure refers to the activity before the events of Tiennenman Square. The government used to support Fuzzy Logic in a big way back then, but now some of the most active people have left for Singapore and Hong Kong.
They've scattered. I guess the Chinese government didn't like that so many of the leaders of that movement turned out to be Pro-Western and as some sort of punishment, they've decreased the funding for Fuzzy Logic, not to zero but much lower than it was before. I haven't been to China to see what's happening since 1985. My understanding is that support is beginning to pick up again and is likely to grow substantially.
There seems to be a close correlation between Fuzzy Logic and Linguistics. So much of what is inherent in Fuzzy Logic relates to the way people think and talk-in other words, their use of natural language. What influence did your early exposure to so many different languages play in shaping these attitudes? (Before Zadeh was twelve, he was having to deal with four different languages-Russian, Azerbaijani, Persian, and English-and three separate scripts-Cyrillic, Arabic, and Latin).
You're right. The Fuzzy Logic model relates closely to linguistics. But I'm not sure learning these languages had a big influence on my thinking. If it did, then only subconsciously. But I do remember one thing that made a very deep impression on me in my youth. That was how different people could wholeheartedly embrace systems-whether political, religious, social, whatever-that were diametrically opposite.
After leaving Azerbaijan as a child, I attended Alborz College, a Presbyterian missionary school in Tehran. Every morning, we had to go to chapel-that was a drastic change from what I had experienced at the Soviet atheistic schools. But from this experience, I grew to be tolerant of many different points of view-Soviet atheism, Protestantism, and Muslim fundamentalism.
At the same time, I'd have to admit that I did gain something from early exposure to all these cultures. It's clear now as I look back. For example, the Soviets placed science and technology on a very high pedestal. They also instilled the belief that you owed something to society. That you shouldn't be self-centered, egoistic, seeking only your own pleasure. That you should focus on what contribution you could make to others.
In Iran, I was deeply influenced by the decency of these American missionaries who ran the school which I attended. It's the sort of decency that one finds in the US if you go to the Midwest or rural areas-away from the big cities. To me, these people were role models-so willing to help others who were not of the same ethnic origin. They weren't nationalistic. They had a mission and they stuck to it. That influenced me deeply.
Then while studying at the University of Tehran, the intellectual climate at that time was deeply influenced by the French culture as most of the professors had been educated in France along the lines of Cartesian tradition. That means they were very precise. So even though you might be thinking about very imprecise things, you had to think about them very precisely. Fuzzy Logic is like that. It's really an exact way of thinking about very ambiguous and obscure things.
(Zadeh pulls an issue of the International Journal of Fuzzy Sets and Systems from his shelves and thumbs through the pages.) You see, pages and pages of very precise mathematical formulas. Fuzzy Logic is not abstract thinking. In reality, it's very concrete. It's built on very precise formulas.
Also from Iran, I learned a certain kind of warmth. People are warm in that part of the world. There's a great depth to their friendships. People tended to be polite. Looking back, I feel it was a very positive experience for me to have grown up in Iran. There's a great respect for knowledge there. Even if an individual didn't have anything by way of material goods, he was highly respected if he were knowledgeable.
And what about the influence from Azerbaijanis?
Obstinacy and tenacity. Not being afraid to get embroiled in controversy. That's very much a Turkish tradition. That's part of my character, too. I can be very stubborn. That's probably been beneficial for the development of Fuzzy Logic.
And Russia? Your mother was Russian, wasn't she?
From the Russians I gained a great respect for knowledge-a broad based knowledge. I read so many books as a child-Dostoyevsky, Chekhov, Tolstoy. Those books put goodness on a pedestal. The heroes of those books were always fighting evil. You'll find the pursuit of goodness on all their pages, except for Dostoyevsky. I had a library of, maybe, 3,000 books since my father had been a journalist. I read Shakespeare in Russian. There was such a respect for knowledge in my youth.
There was another factor that deeply influenced my early life, too. I grew up as an only child. I remember I used to have a sign over my desk that read, "Alone" (in Russian). I was alone much of the time and that enabled me never to feel pressure to be like others or do what others did. I never felt I had to go along with the crowd. I've always felt separated enough so I could go my individual way. It's always been that way.
You must be quite pleased to see how the field of Fuzzy Logic has expanded so much in these past 25 years, especially since the late 1980s.
I don't feel any different now than I did 15-20 years ago. Somedays, I'm pleased with the progress; sometimes not. C'est la vie! But basically, I don't have an egocentric view of myself. I never felt that I was an important person. I feel ordinary like anybody else. I have no feeling of importance. No such feeling whatsoever. I don't attach very much importance to who I am.
Looking back on the development of Fuzzy Logic and its application, what would you have done differently to promote your ideas?
I've made no effort to go after scientists or companies to show them how my theories could be applicable to their work. I don't promote this thing. If people want to do something, that's fine. If they don't, that's fine, too.
What kinds of applications have you been excited to see develop?
I can't say that anything has been "exciting". Rather, I would choose the word "interesting". Not too long ago, the Chinese University of Hong Kong conducted a survey to determine which consumer products were using Fuzzy Logic. The result was a thick report, some 150-200 pages long-washing machines, camcorders, microwave ovens, etc. What interested me wasn't the particular applications so much as the breadth of applications-so many products were incorporating Fuzzy Logic.
What would be the ultimate application of Fuzzy Logic?
I can't say that I have much time to think much about such things because I have to focus on more immediate deadlines-like this speech I'm giving in New York to a Semantics group on November 4th (one week from now). I tend to think about things closer to the present. I have to do quite a bit of thinking about those things; if I didn't, I'd soon sound like a broken record and they wouldn't invite me to these conferences. I always feel like I have to say something different than what I've said before.
What's the future for Fuzzy Logic?
In general, I think the future will involve fuzzy logic, neural networks and genetic algorithms. I lump all these under the rubric of "soft computing". I'd encourage people who have the inclination and ability to become competent in all three of these areas. Eventually, I believe that Fuzzy Logic will have a wide-ranging impact once it is understood how widely the theory can be applied. Of course, this is just my opinion; only time will tell.