Welcome to the last installment of Traveling in Cyberspace. Mike Coovert wanted to make this issue of TIP a special one, given that it would be the last under his editorship. He challenged us to come up with a very special topic on how technology will change the world of work in the twenty-first century, and we rose to the challenge. We wanted to interview a visionary, someone who has had, and will continue to have, a major impact on technology. We could think of no one better than Donald Norman, who was kind enough to grant us an interview.

Dr. Norman has been a leader in the fields of cognitive science and human-computer interaction for many years. He served on the faculty of the Department of Cognitive Science at the University of California at San Diego from 1981 to the present, and is currently Professor Emeritus. From 1993 to 1997 he served as Apple Fellow at Apple Computer, Inc., and as Vice-President of Apple Research Laboratories from 1995 to 1997. Currently, he serves as head of the Appliance Design Center for Hewlett-Packard. He is the author of multitudes of publications and books. One of his more popular books is The Design of Everyday Things, published by Doubleday. (This is one of our favorite books of any kind, and is a MUST read for anyone who thinks they are too inept to program a VCR! Hint: It’s not your fault, it’s badly designed!).

Because Dr. Norman’s vita is actually longer than this column, we’ll point you to his Web site for more information: http://cogsci.ucsd.edu/ ~norman/default.html

What doesn’t come across on these written pages is Dr. Norman’s great sense of humor. And without further ado, here is our interview with Donald Norman:

*****

P.C. I write a column with Jason Weiss, my graduate research assistant. It’s called Traveling in Cyberspace. It’s written for the national publication of the Society for Industrial and Organizational Psychology. We’ve got a special edition coming out in April, and what I wanted to do is ask you a few questions about some of your philosophies that you developed over the years, particularly in terms of computers, humans, and the world of work.

D.N. OK. Hi Jason.

R.J.W. Hello.

P.C. First, I wanted to give you the opportunity to tell us a little bit about your new book called Information Appliances that is coming out from MIT Press later this year. Could you give us an overview of that?

D.N. Sure. Actually, the title has changed. The title was originally going to be Taming Technology. It’s what it is about. I then changed it to Information Appliances, which is my suggested solution. What it’s really about is making computers invisible. The new title is The Invisible Computer.

P.C. I like that! Could you talk a little bit about what you discuss in your book?

D.N. Mainly I believe that today’s personal computer is just fundamentally the wrong approach. It’s a machine that is developed by technology companies, by technologists, for technologists. Yet the majority of the public are not technologists, and the public wants consumer goods. In the early days of any technology, when a technology isn’t really quite up to par, products are technology driven, and they are driven by engineers. Basically you could do no wrong, and with each new release of the product you have more and more technology, and that’s what customers clamor for.

With time, as technologies mature they move into the consumer phase. Now what consumers really want is convenience. One of my favorite examples is that you go to the kitchen to use an eggbeater. The name of the device in fact tells you what it is for and it does not name the technology. It is not named an electric rotor with hooks.

P.C. [Laughs] That’s true.

D.N. So we still go to use our personal computer, which is wrong; it’s not really very personal and we don’t want to do computing.

P.C. That’s right. In fact, that’s one of the things that I’ve read that you’ve said. The word "personal computer" is a really poor term for what we call the PC.

D.N. Right, exactly.

P.C. Well, what would you have called it?

D.N. Well, what I would prefer to call it in the future is by the name of the task it does. So when I say I would like to see information appliances what I want to see is the computer disappear into the device. For example, if you take a musical keyboard or a drum machine or an electric guitar, you don’t think of these as computers, and you don’t think of them as information appliances, but they are. They’re computers, they’re information appliances, but they’re very well specified by the task that they are doing. The whole physical shape and form is specialized for the task. And they have a very powerful standardized information protocol so that any one of these can talk to any one of the others. It’s called MIDI. And that’s the paradigm that I would like to see.

A calculator is another good example of an information appliance. Even though a computer really does a fine job of calculation, we quite often prefer calculators for lots of reasons. One factor that is very nice is that it is very easy to enter the equations or the numbers. We have specialized calculators, whether it be a normal arithmetic one, or a scientific one, or a financial one, and more important, you can take the calculator to where you do the work. Today, we have to take our work to where the computer is located. So the only problem with the calculator is that it is self-contained and independent so that you can’t connect it to anything else and take advantage of, say, some other device that might spit out numbers which would be just perfect for the calculator to take as input.

P.C. That is interesting. As an industrial psychologist one of the things with which we are concerned is training, and too often we find it is very difficult to get people trained for a job, or a particular device, because of the resources that are available. For example getting a trainer to a particular location and getting everybody involved together at one time. One of the things that has been touted about the Web is that it has the ability to provide training to workers and educating people of all ages anywhere, anytime, just-in-time training to mass audiences. In your opinion, is this a "pie in the sky" idea, or do you think that could be reality in the future?

D.N. That’s a very complex question, so let me take it apart. The normal people would say "hooray!" for the Internet and how we will dramatically revolutionize training. That’s not a pie in the sky, that’s a pie in the face. So basically, what is the worst way of teaching somebody? Answer: the lecture. So we take the worst way of teaching somebody and actually make it even worse. Then instead of lectures, we put it in text and little cam videos and now we distribute it over the Internet. This does not enhance education. Just-in-time learning actually uses a very, very different concept. The proper just-in-time learning is exactly the minimalist training that Jack Carroll talks about. You know his book about minimalist training?

P.C. I’m familiar with it, yes.

D.N. What Carroll discovered was that all these big manuals were worthless because people looked at them and said "Yaaa! Why do I want to read this? I just want to know what to do." And so what he proposed and showed and what really worked was that you give people the absolute minimal instruction and let them get started on their task. And they’re really happy because they’re actually accomplishing something. And then as they’re doing their task they run into trouble. That’s when they’re ready for just the right amount of information to tell them how to proceed. That’s what just-in-time learning is about. It’s not bombarding you with lots of information. But it’s giving just what you need at the time that you need it. And this is not what most of the people are talking about when they talk about Internet training. I was just at a conference and I saw some novel schemes, Internet training, and these people confuse cuteness, for example, with education.

P.C. Right. I’ve seen that. More glitz than substance.

D.N. Right. I think the Internet can revolutionize things, but the traditional way it’s being done, I’m not a fan of. The one institution that I think has done an excellent job at learning at a distance and using the Internet is the Open University in Great Britain.

P.C. I’ve heard of that, but I have no experience with them

D.H. Check them out on the Web. They have something called Knowledge Management Institute. I don’t remember the URL, it might be KMI.

P.C. Oh, I’m sure I can find that. Is that part of the Open University?

D.N. Yes. It might be KMI at OU. Who knows what it would be? Whatever the complex stuff might be probably .ac.uk or something.

P.C. One of the things that you have prominently on your Web site is the motto from the 1933 Chicago World’s Fair that says "Science finds, industry applies, and man conforms."

D.N. Right.

P.C. And then you have your own person-centered motto, which I liked very much which is "People propose, science studies, and technology conforms." Can you expand on that? That is, on your person-centered motto for the twenty-first century, in particular in terms on how it might apply to the workplace, and how work will change because of technology?

D.N. Actually, what I’ve taken is from my book Things That Make Us Smart where I do go into this in more detail. And actually in my new book, The Invisible Computer, I discuss what happened with the mechanization of work so what happened around the turn of the century as we introduce so- called scientific work practices, Taylor’s work on scientific work practices. And for that matter, Henry Ford’s work on the assembly line, and earlier than that, the work on the meat packing factories where they did what might be called the "disassembly" line. The whole philosophy there was to try to take work apart into small segments to treat humans as machines, and then give each person very tiny segments that they could do repeatedly. Taylor, for example, thought of his laborers as brute laborers and didn’t think they were capable of thought, and on top of that, didn’t think it was appropriate for them to think while in the workplace. Because after all, if you stopped to think, it would slow you up. So his notion was that you should just decompose the work into small segments and that required no thought and you could do efficiently for the whole day. What he failed to understand was people. That’s just not a very high morale situation. It dehumanizes the workforce, and it leads to alienation of the workers from the so-called "superior" managers who are doing the decomposition and who are doing the thinking. I think the current status of the workforce in many ways had its roots around the turn of the century as we try to do scientific analysis of work habits. It didn’t take into account human psychology.

What is happening today, I think, is that we’re better able to understand that you need motivated workers. You want workers to feel that they have made a sizeable contribution to the device that they are producing, to the product. And that makes a dramatic difference. And even if they are a little bit slower at each operation, by the end of the day, they are much more efficient. So we see some of the harm that happens to long distance information operators. Psychologists determined that if you can save a second or two off from each telephone call, you can save tens of millions of dollars for the telephone company per year. And as a result it is a horrible job—you just sit there, and you don’t even say "hello" to the customers. The "hello" to the customers is now done today by a computer-synthesized voice, and they only get to hear what your question is, and then they turn you over to an automated system that gives you the answer. You may have noted that you get to talk to the operator for only a few seconds and then that person is on the next call, and the next call, and it’s a horrible position, a horrible job to have.

P.C. One of the things that I’ve found interesting in the last few years is the issue of telecommuting. Business teams no longer have to be situated in a particular physical or even geographic location. Business teams now, especially for larger organizations, can be situated throughout the world. This appears to be totally technology-driven, you know, the ability to communicate with each other in various ways, and also an increase in the bandwidth has created that flexibility to not have to be situated in a particular place or even a particular time.

D.N. Well, some of it you might think of as technology but some of it is out of necessity. Yes, technologies have allowed companies to become international. Before the days of rapid travel and rapid communication, companies could not actually service an area much larger than their homebase. And before the advent of the telephone, the company had to have its sales in the same place as the manufacturing plant. The telephone made a big difference in that. So did the telegraphs and before that the trains, and today, of course, the more modern communication methods. As for international situations, for example, I’m working with a group in Bristol, England. I can’t afford to go to Bristol, England every week. So working at a distance becomes absolutely essential. I used to have people that worked for me that lived in Minnesota, so new technology allows this. So I’m not sure it’s technology driven. I think it’s actually need driven.

A very interesting book that I’ve been reading that is called the Death of Distance, which is arguing that although distance is now disappearing, time zones remain.

P.C. [Laughs] That’s true.

D.N. In fact, it’s time zones and language that determine more and more how people collaborate with each other. So I’m having great trouble actually working with my group in England, because unless I can contact them before ten o’clock in the morning they’re not at work any more. And I tend to be a late worker; I tend to get to work at ten. So the window is almost non-existent. I find the real problems are time zones. Now we have found in doing software development we can take a chance at the time zone. We have our software group in India start to work on our project, and when they go home for the evening they send it off to California, who works on the project, and when they go home for the evening they send it off to England, who works on the project. So what you have is 24-hour-a-day work on the same project.

P.C. So one team would work on a piece of software and send that to another team which would continue on?

D.N. Not all projects allow that kind of handoff. But there are certain things, especially in the testing phase, where you can actually do as much testing as you can in 8 hours, and then pass it on to the next.

P.C. I’ve got two more questions, I know you’re very busy, Dr. Norman. One of them deals with something that was in Chapter 6 of Taming the Technology—what your book was formerly called.

D.N. At this point, I don’t remember, the book has been changed so much. I promised myself this weekend to put up a more accurate version.

P.C. What it referred to was activity-based appliances. If you could to describe how you got started at Apple with activity-based computing, in particular what ramifications you think that has for selection of workers. Because now we start with a job that has been designed, and we have hardware and software and other implements that are used to complete tasks. As industrial psychologists, what we do is try to match the person with the task. But it appeared to me that some of things that you were saying about activity-based computing could actually allow you to be more flexible in selecting workers. What you would do is engineer and design the computing devices to adapt to the individual. Am I off-base there?

D.N. Actually we didn’t emphasize adaptation there. We emphasized trying to do good task analysis, if you will, and designing the system around the full activity that is to be performed. Instead of trying to optimize some local thing like word processing, people don’t want a word processor. They are writing letters, or papers, or memos, or something. People don’t write checks, they’re doing financial transactions, of which writing the paper or writing the checks is one small component. So the argument is, if you can support the whole activity, the tool is much better fitted, much better in aiding you to accomplish this task. My opinion is that it is pretty simple and is a pretty straightforward idea.

P.C. Did you think that some of these systems that support work are going to involve artificial intelligence?

D.N. Artificial intelligence is already in active use. A lot of intelligence search engines use techniques from AI. But I certainly don’t believe in a system that tries to read your mind and try to get your attention.

P.C. Oh, no. Right.

D.N. In those senses no. In the sense though of doing intelligent searches and finding things that are say, similar to the document that you are looking at, then yes, then those have already proven to be effective.

P.C. I just have one more question. My GUI Design class would appreciate your answer on this. Something that you said in the same chapter when you were talking about what’s wrong with the GUI’s today. What you said was that they "just don’t scale" and you wrote "making everything visible is great when you only have 20 things but not 20,000." My question to you is what do you think that some of the interfaces of the twenty-first century will look like; how different will they be than some of the current interfaces we use?

D.N. Actually I would hope that the best interfaces are the ones that you don’t notice at all. Think about what the electric guitar is. If you actually think about it, the electric guitar is a pure interface mechanism. You play music, you play electric guitar—talk to somebody who uses one. What it is, is a wonderful set of strings and frets, and you can basically use this to control the sound. But the kinds of sound that come out of them is determined by the sound generator. It has nothing to do with guitars anymore, and that’s a clever interface.

What I expect to see is more and more interfaces that are tailored for the task that you are trying to do. I just saw a demonstration of a gesture interface where you just move your hand and point to the object. You can play games as to where you see the ball coming at you, you lift up your hand and you swat it. I can imagine in doing gardening, my gardening tool might have a probe that you stick in the ground and made with a little camera that I aim at the plants so that the system can measure ground acidity and moisture, and identify the plant, photograph it, and tell me what I should be doing. What I expect to see is the interface tailored to the task far better than what we do today. So today we sit down with the mouse and the keyboard, which isn’t relevant to gardening or relevant to making music. So that is what I expect to see. As the computers become invisible and we are embedded in task-specific devices, the interface will be appropriate for whatever work we’re doing.

R.J.W. Dr. Norman, if I could ask you a question myself. When you talk about having particular tools such as the calculator, and that’s just a calculator. The computer can do a number of things that have no relation to each other, such as word processing and surfing the Internet. So you mentioned before you don’t like the term "personal computer," it’s not a good term. What term would you use for such a multipurpose tool?

D.N. Why do I want a multipurpose tool?

P.C. [Laughs] Good question.

R.J.W. Well, to save on space, I guess. Instead of having a word processor and something that sits right beside it that surfs the Web.

D.N. You’re missing the point. How many devices do you have already in your house? You have a huge number of devices in your house. The point is, you put them where you need them. You don’t stick them all on top of each other. So I don’t want a word processor by the way, I want a letter writer or I want a report writer or a paper writer. And I will put that where I want to write my papers or my letters. And I won’t have a Web browser, I suspect. I’ll have many of them, so there will be a Web browser built into the TV set and I’ll use that for entertainment, and checking out advertisements that I have seen on the TV shows that I watch. I’m sure my children will have a homework machine, and they will browse the Web probably provided to them by the school. I’ll have a financial center where I pay my bills, and my income tax, and maybe my investments. I’ll browse the Web, probably provided by my financial institution to learn more about my investments.

So how to make everything into one device misses the point. I want to choose to work in the different locations in my house, and there I want a tool that is appropriate. I just saw a very nice device at a communication center, the telephone. It’s got a telephone, voice mail, and e-mail, and does Web browsing. I would have that in my kitchen but I wouldn’t do full Web browsing. I might use it to look up the weather, or I might connect to my son’s school connection so that I could see when the soccer game was, or when the plays are. So I would expect to see these scattered all through the house, much like all of our normal devices are scattered throughout the house. You have many telephones throughout your house which you put them where they are convenient. Does that answer your question?

R.J.W. Yes.

P.C. That surely did. Dr. Norman, our time has run out. I want to thank you very much for sharing your thoughts and philosophies with us.

D.N. Fine, thank you very much. Bye.