SUBSCRIBE NOW!

Vol. 11 No. 9 — October 2003

Ardito: We’re asking everyone we talk with about the formative experiences of their very early life.

Cuadra: I was born in San Francisco and went to Lowell High School, which was a kind of prep school for the Bay area. I left high school after 3 years and went into the Navy. I took Armed Forces Institute courses and got my diploma while I was still in the Navy. In 1946, I enrolled in the University of California, at Berkeley. That was the University of California at the time.

Bjørner: You went into the Navy in 1944, during WW II?

Cuadra: Yes. I served overseas. I was a radioman second class, copying Morse code: listening all day and typing code. I went to the Pacific, to Manus Island, part of the Admiralty Islands, off New Guinea. There was an officer there who created an off-duty educational program that came to be called the College of the Admiralties. The officer directed it, but I was put in charge, living in the building, arranging courses, and finding volunteers to teach English, photography, bookkeeping, etc. There wasn’t much to do on that island, so people were perfectly willing to teach classes. We awarded diplomas and helped many move along in their education.

Education and Entrepreneurship

Cuadra: After I was discharged, I studied general psychology at Berkeley. For some reason, I took Psych 1A and 1B at the same time. Apparently, I took very good notes; people would sit next to me or borrow my notes before class. I discovered there was something called Fybate Notes. It was a business where good note-takers would audit classes and provide their notes, to be printed and sold to students who took poor notes or didn’t go to class. I bought a mimeograph machine, printed my notes for Psych 1B, and sold them for pennies to other students. The professor got wind of this and offered me a job as a teaching assistant. He wanted to keep me from selling notes, so he gave me a job. My entrepreneurial business stopped dead, but I worked for that professor for several years.

Ardito: Were you on the GI Bill at Berkeley?

Cuadra: Yes. I think I did my undergraduate work in 3 or 3-1/2 years. Then I decided to work toward a Ph.D. in clinical psychology. Ph.D.s were not associated tightly with any specific area then — a Ph.D. was a Ph.D. But there were several possible specialties, and the one I chose was clinical psychology. My interests were always broad. The clinicians thought I was an experimental psychologist, and the experimental psychologists thought I was a shrink, because I wasn’t purely one or the other. I took experimental psychology, animal psychology, statistics, and subjects that aren’t clinical, as well as the clinical areas. I got my doctorate in 1953.

The last 3 years included internships. I spent some time at the University of California, Berkeley, Mental Hygiene Clinic; then at the Oakland general medical hospital; and finally, the Palo Alto Veterans Administration psychiatric hospital. They sent us to different places to get exposed to different kinds of patients. After finishing, I wanted to get a job in California but couldn’t — at the University anyway — because the policy was to go someplace else, get cross-fertilized, and then come back. I took a job in Downey, Illinois, at the VA Hospital. I was in the clinical psychology department there — doing therapy, testing, and training new trainees — for about 3 years.

Ardito: Is that the first time you were outside of California?

Cuadra: Yes, except for the service. I didn’t love it. I really wanted to be in California.

The RAND Years

Cuadra: In 1956, RAND Corporation was recruiting widely, and a recruiter came to Illinois. RAND had invented a training concept that involved cutting off live communications to radar sites and substituting computer-generated inputs. That way, site personnel who never saw many airplane blips could practice dealing with massive attacks by hostile aircraft.

Bjørner: Those were early simulations!

Cuadra: Exactly. For two hours, with live radar cut off, fake inputs would come in. People on the phone would pretend to be at the airfield where the fighter warplanes were stationed. They would say, “Scramble two fighters,” simulating sending them off. This was a very well-done simulation. At the end of 2 hours, there would be debriefing and feedback. That was the invention, and the Air Force contracted with RAND to install the simulation program in every radar site in the country. RAND had to hire people to train the Air Force. I was one of the people RAND hired. It was a secret program. The person who recruited me, Dr. Elias H. Porter, was very persuasive. He said, “I can’t tell you what you’ll do, but you’ll really like it.” The combination of his personality and the fact that it was in California was enough to let me walk away from clinical psychology forever.

Ardito: Did they send you to different training sites?

Cuadra: No, I worked for a while doing the training. I went on only one of the 6-week trips. Then I was promoted to be one of the recruiters and trainers, so I hired other people and sent them out. The program started with the manual air defense system. Then the Air Force automated the radar system, calling it SAGE (Semi-Automatic Ground Environment). RAND needed people to train the operators of SAGE as well, so the program I was in grew to prepare people to go on site and train the Air Force in the use of this computer-based air defense system.

Early Computer Experience

Bjørner: That was your first experience with computers, then?

Cuadra: That’s right. I’d never seen a computer before. The first one I saw was the SAGE installation in Santa Monica at System Development Corporation, in a big building with no windows. It was an IBM computer called AN/FSQ-7, and it took up about 10 times the size of a 15 x 20 foot conference room. You had to walk between the components of the computer. There was a memory unit the width of my outstretched arms and about 5 feet tall. Another unit was drums — there were no disks in those days. The CPU was still another unit. The system took up the whole room, and there were two! — it was duplexed for reliability.

This was a firstgeneration computer: vacuum tubes, drums, no disks. The memory was only 64,000 bytes (64 K). It was interactive, supporting more than a hundred consoles, so each activity had to be completed on a scheduled basis. There wasn’t enough memory to do everything at once and support all these consoles, so they just shuffled through processes, with the most important jobs — such as identifying aircraft — done most frequently. Many of the people I hired were Ph.D. psychologists who became team leaders when they went out to sites. There were four SDC people to a site. They all had to learn about computers, so we had computer, air defense, training, and how-to-do-your-job classes, all of which I was in charge of arranging.

Ardito: Was SDC one of the sites that you were working on for RAND?

Cuadra: No, RAND invented the program, and the Air Force decided to implement it. The Air Force hired RAND, but RAND didn’t want to run the program, because its mission in life was to give advice to the Air Force. RAND had never sold services other than providing advice. A division within RAND, called System Development Division, was set up to run this program. Later, RAND spun the Division off, setting it adrift, totally disconnected from RAND, with its own contracts with the Air Force. In 1957, when it was split off from RAND, I was moved to System Development Corporation.

Bjørner: How did you get trained on using the computer?

Cuadra: There were training programs, but our trainers were not expected to do anything with the real operational computer. Our team would hand the Air Force a tape; the tape would be loaded into the computer; and this would do what the old program did — cut off all external inputs and substitute inputs that were computer-generated. The machinery to make these inputs was very complicated; when an airplane passed out of one sector, it had to disappear from that radar screen and appear on the screen of the next site. All these inputs had to occur as though they were really happening at more than one site. Our people didn’t learn how to be air defense experts; they simply learned how to conduct the exercise. They didn’t even do the debriefing; they just helped the Air Force officers and staff debrief. We told the Air Force how to run the program.

Ardito: So you weren’t doing any programming yourself?

Cuadra: No. Never did. I’m not a programmer — couldn’t write a program to save my life. This project went on for 2 or 3 years. Then I had a sequence of jobs, and, at some point, I wound up running the library and documentation systems department. This department did research for various government agencies, primarily the Air Force.

Information Science and Technology

Ardito: Did SDC develop the ORBIT software for the Air Force?

Cuadra: Yes — the original version — but I was not involved in the early stages of that development. There was a lot of information-related research going on. Government contractors got a certain amount of money for independent research, and SDC staff were doing all kinds of things — document association, computer database retrieval — many projects. Other people were developing retrieval programs long before I was associated with them.

Ardito: How did you get involved in the information science side?

Cuadra: I don’t remember what unit I was in, but in 1962, I wrote a paper entitled, “Who is Fred Dutton?” A silly name, but its point was that SDC needed a way to tell what prospects SDC was marketing to. There were some embarrassing situations, where people from one office of SDC were trying to sell something to an Air Force unit, not knowing that another group was also trying to sell to the same people. My memo said that there ought to be a contact file, and I described the kind of information that should be in the file. I don’t know if anything was said about computers. This is the earliest event I recall that shows my interest in information processing.

In 1963, I went to a meeting of the American Documentation Institute (ADI). I have no idea how I heard about these meetings. I probably thought that ADI was a place where people who were interested in information and documents hung out and maybe I would learn something. The meeting was very mysterious to me; people spoke a language that I wasn’t familiar with. I got interested in the meeting itself and did a survey — an evaluation — of the 1963 meeting. I sent out questionnaires and reported on the results. There’s a part of me that always tries to figure out what I’ve just done and how it can be done better. I was trying to get a grasp of what was happening at the meeting, without being an expert in the language and the issues.

Bjørner: This was a national meeting?

Cuadra: Yes, and what I wrote was a technical report, published by SDC, because they paid for my attendance.

Then I started wondering what I should read to learn about this field. I wrote to at least 50 persons — I don’t know how I got the names — and said, “I’m trying to imagine what someone from Mars would find useful if they want to learn what the documentation field is about. I would appreciate your telling me what you consider to be your most significant published contributions, and why, so we can identify starting places for this Martian.” Many of the leading lights — Allen Kent, Robert Fairthorne, Charlie Bourne — wrote back with very thoughtful replies. In 1964, I published the responses in a document called, “Identifying Key Contributions to Information Science.” By reading the various publications and reports, I began to learn about the field.

The Annual Review

Cuadra: I started mentioning to anyone I talked to that there ought to be an annual review in the field — to distill the information, to present what is most important, and to show progress. I was familiar with annual reviews from the field of psychology and found them very helpful; they exist in many other fields. I said this for 2 3 years or so, and finally a colleague at SDC remarked that the only way it would ever start is if I were to do it. I said, “I don’t think I’m the one to do this — I’m not a professional in the documentation field.” I’d seen enough to know that I wasn’t!

About this time, Helen Brownson, at the National Science Foundation (NSF), and I had some correspondence. I think she had heard my message, and she expressed interest in my undertaking an annual review. I wrote to Helen, saying, “I still think I’m not the right person for this, but my company said they would contribute my time to get it off the ground.” So NSF provided the funding. Helen insisted that I be the editor. The ADI wasn’t very hot for that idea — I wasn’t one of their leading lights.

Bjørner: Were you a member of the American Documentation Institute at this point?

Cuadra: Yes, I probably had joined. But there already was a person whose position was editor; that was Art Elias. He didn’t like this idea of a new editor position at all and fought fiercely against it. I think several people fought against the name I wanted to give the new publication, which was The Annual Review of Information Science and Technology. They didn’t like the “Information Science,” and they didn’t like the “and Technology.” But Helen said, in effect, “You either accept Carlos as the editor or there’s no money.” So they held their noses and said, “OK.” The first volume was published in 1966, by ADI. There was enough money for two volumes.

Ardito: Did Art Elias ever talk to you again?

Cuadra: Oh, yes. Art put a wonderful curse on me at a meeting. He said, “May the hairs on your head turn inward, seize upon the follicles of your brain, and erase your memory.”

Ardito: What a curse! The title of your Annual Review seems to have influenced the changes made to the name of the organization. The American Documentation Institute subsequently changed its name to the American Society for Information Science, and now (years later) it has added “and Technology” to the name.

Cuadra: Yes. I came to the field kind of sideways, not from within. The word “documentation” didn’t mean anything profound to me, but it had been the name of the Society for so long that no one wanted to change it. In Chapter One of the first Annual Review, there’s a discussion of the nomenclature.

Bjørner: It does sound as though you were likely responsible for the name change of ASIS, because of the title you put on the Annual Review.

Cuadra: Yes, I could lay claim to that. Anyway, I was learning so much, and there is no faster, better way to learn anything than by editing. I did what any editor does. I tried to make sure that what was said was in English and that it was clear. I would pick at words and contradictions between something said on page one and something said on page 17; then I would send back detailed letters to the authors with an apology, saying, “You get to decide what to pay attention to, but here are the problems I see.” The manuscript would be marked up with notes and corrections. Everyone, I thought, took it well — at least they ended up making better chapters. I told them, “I want to make sure you say what you mean and mean what you say, so I’ve tried to make this better, but it’s your chapter; you decide.” The authors generally accepted that.

That’s how I learned. I didn’t remember everything I read, but I learned something from everyone. The author of one chapter didn’t deliver; it was a chapter on computers, in Volume 1. I assembled my staff at SDC, talked to anyone else I could find, and we wrote the chapter. This is the only chapter that was not associated with individual names. I put it together but said it was by “Annual Review Staff,” because I got all the input from other people. I didn’t know that much about computers at the time. We needed that chapter because it had been promised in the outline and we didn’t want to disgrace ourselves. So, in effect, by editing material, I learned about things that I didn’t do as part of my daily job.

I documented the Annual Review procedures. I built a how-to-do-it manual that could tell Martha Williams, the next editor, everything I knew about the process to make it easier. I tend to be a systematizer, because I always do enough things wrong that I want to figure out what can be done better the next time around.

What I eventually realized was that I had been interested in words for a long time, starting with my first work in psychology, where I cared about the accuracy of communication. To the extent that I’ve helped drive the online business, it’s really as a communicator, not as a programmer, nor as a technician. It kind of all fits together when I look back 30 years or so.

SDC and ORBIT

Bjørner: Tell us about the environment at SDC, how this was different from what Roger Summit had at Dialog, and why you left SDC.

Cuadra: SDC was a great experience and it was wonderful to be involved in the invention of online. I was the chief agitator for it, but I had a lot of help in creating SDC Search Service, most particularly from Judy Wanger, with whom I worked for 10 years at SDC and nearly 25 years now at Cuadra Associates. Two programmers who wrote ORBIT work for me now, Don Blankenship and Bob Burket. The programmer who wrote the accounting program was my son Neil, who also was the one that saved us by writing the first data conversion program — for the CHEMCON database.

There are two reasons why I left SDC. The fact was that the cost structure at SDC was very different from that at Dialog. We had a computer center that sold its time to the government, practically 100 percent to the Air Force. There were established rates for computer time. When we started the business, the computer center decided that we had to pay the same rate as everyone else. The fact that this was a company-funded product and effort, that we were trying to invent something and start a new business, seemed to be irrelevant to the person who ran the computer center. Maybe there was a policy level above him that affirmed that practice.

The computer center bought the latest equipment and software, because the Air Force was supporting it, needed it, and could afford it. They bought high-powered computers as soon as they came out; they bought the fastest, newest IBM disk drives. All this was expensive.

In contrast, what was happening at Dialog is that Roger Summit, clever person that he is, went out and bought used IBM data cells, which were very slow, very unreliable, and dirt cheap. He bought up every one he could find. So while I was paying high-end retail at SDC, he was getting almost free disk storage. Not only that, the Dialog response time was so slow that, with the connect hour charge, Roger was making two or three times as much money as we were, because our system was stunningly fast.

I saw this once, when I was demonstrating in Mexico. I watched someone searching Dialog, and it took forever; the person didn’t move, didn’t blink. My stomach was churning, because it was so slow. I didn’t say anything but I thought to myself, “This is horrible, why isn’t she screaming?” But users just accepted it; that was the way it was. So, Dialog made a ton of money with this slow, cheap stuff, and the decks were stacked from the beginning.

Connect Hour Pricing

Ardito: But in the literature of the history of online and biographies of key people, nearly all the articles mention that you and Roger Summit got together to come up with connect-time pricing.

Cuadra: Yes. That was the only way to make it intelligible to our target audience of librarians and information people. You couldn’t talk computer cycles; users didn’t know what a cycle was. I didn’t know what a cycle was or how much it cost. Users didn’t know how big the database was, so we couldn’t charge for disk storage. There had to be some metric that was understandable. Time was one of those. If you dawdle and spend time there and search a thousand things and take 2 hours, you pay more than if you zip in and zip out.

Ardito: But was the pricing decided separately? Did you come up with it, and Roger came up with it, but separately?

Cuadra: Yes, there was never any discussion. It may be like the airlines, where someone raises his price and everyone else says, “Oh, let’s do that, too.” I don’t know who did what. I never considered any other model.

Ardito: When did you realize that ORBIT was so much faster than Dialog, and also that it seemed as though the users were not aware of that? Or didn’t care.

Cuadra: Right from the beginning. Users just accepted it. Unless they had access to both systems, they wouldn’t know the difference. They would look at the two price sheets, and the prices were more or less the same. You wouldn’t know there was a difference unless you actually did the same search on both systems and compared them. Some people did.

Bjørner: That was one of the differences between SDC and Dialog, then. You were paying retail prices and you were also paying for state-of-the-art technology.

Computer Center Trials and Tribulations

Cuadra: I complained constantly, and my relationship with the guy that ran the computer center was very bad. He didn’t like our use of the computer, because it was different from the other uses; he really didn’t like me; he didn’t like what we were doing. Whenever we had to reload something, we had to prove that he had made a mistake. Sometimes the operators would mount tape three before tape two, and screw up the database so something had to be done. I had to prove him wrong to his face in order to make the change. The climate was terrible, and the fault was above us both; someone higher up should have said, “Here are the rules; here’s how we’re going to play this.”

At one point, SDC took my programmers — who understood data, what a field was, and what an index was — and assigned them to the computer center to work under the professional computer people. The professional computer people would say that a given operation — to create a data conversion program, for example — would take 7 or 8 weeks. I knew it needn’t take that long, because the same programmers, while they worked for me, could often do it in 1 week. But the computer center would make us wait 6 more weeks before delivering the program. That was the kind of activity that was tolerated by our management.

I kept trying to work around it. I finally got the programmers back working for me again. I have to tell a funny story about Bob Burket and Don Blankenship. We were working on the Derwent World Patent Index database, which required us to insert space for equivalents linking the original patent to each country’s number. It was a very large database, and we discovered that when the database got to a certain size, ORBIT could not get over to the next set of disks; ORBIT simply could not expand. There we were in the middle of loading the new database, with Monty Hyams sending nasty letters to me every week, and we discover that we’ve hit this brick wall! There were probably 15 people in the computer center who assembled in a conference room and solemnly told me they had analyzed the problem, and there were two possibilities: one was to re-program to definitely solve the problem, which would take 6 months; the alternative would take 3 months, but it might not work. What did I want to do? I thought that 6 months was not an option; we’d be dead. Three months was not really an option either, but it was the better of the two. So I accepted their 3-months proposal.

I walked out of that meeting and ran into Don and Bob. They saw my face and asked, “What’s the matter?” I explained, and they said, “That’s a one-card change. That’s one line of code.” They explained that all that was necessary was to change one instruction on one punched card. That would fix it. So I called a meeting with the computing center, and Don and Bob explained how to fix this. And the computer center staff said, “We’re going to have to consider this.” They had to consider whether they would accept it! They ended up agreeing that they would let Don or Bob fix it, but it would take a whole day because they needed to test it and make sure it was right. I learned then the difference between a programmer who knows the code and programmers who don’t. I also saw further the malevolence in our computer department. They could have talked to Don or Bob themselves before concluding that the job would take 6 months.

The computer center never adjusted to what our business was. They were supposed to turn ORBIT on at midnight; ideally they should have turned it on at a quarter to midnight. Sometimes they would forget. So I got in the habit of staying up till midnight, telephoning to check whether they had turned ORBIT on. I remember one night when I called in — well, I won’t tell you what he said, but I’ll translate it to “Oh, hell.” It happened regularly. I have a younger son who is a night owl. After awhile, I commissioned him to be the one to check every night to see if the computer center had turned ORBIT on.

Bjørner: It’s really obvious now why you built STAR — so you could be free of programmers.

Cuadra: That’s right. Exactly. Having to beg people to do things is something I wouldn’t wish on anybody.

After several years, I got tired and I got bored. Getting the next database, loading the next database, getting the description out — all of that is like turning a crank, very repetitive. The pain of dealing with the computer center and not making enough money to please the company just got to be too much. I don’t think we needed to make a lot of money, but management did not want ORBIT to be a drain.

Bjørner: Were you losing money or was it cost recovery?

Cuadra: I can’t remember. We were probably not bringing in as much as we spent, but it depends on how you do the bookkeeping. The big number was computer resources. SDC was used to contracts for $30 to 40 million with the Air Force; they would bill $100,000 a month. I was billing one organization $27.42 a month, and another organization $200.10. There were thousands of these.

We did our own billing, because SDC was not geared to deal with large numbers of customers and small amounts of money. My son Neil built a billing program so we could produce a report from the computer, automatically generate our own invoices, send them out, collect the money, and hand the payments to the company.

Leaving SDC and ORBIT

Cuadra: The crowning blow came when SDC sent an accountant to meet with me to learn the details of running Search Service. I think they had planned to eliminate me and needed to figure out how to run the business. He met with me for 4 hours a day to find out everything — how you pick the databases, how you load them, how you get the agreement, how you sell. I explained that you go to conferences with a telephone and modem and you dial into the computer to demonstrate the service. I talked with him, explaining everything, for 2 months. Then one day, he came down with a bill for telephone service and asked, “What’s this telephone bill for?” We had rented a phone line for one of our regular exhibits. After 2 months, he still did not understand that the purpose of the telephone is to contact the computer in California. I decided that this was the end. I felt like I was dealing with idiots.

About this time, SDC management decided that I needed an assistant department manager and they identified a person that I considered something of a lightweight. Since this new position would have put him over Judy Wanger, I decided, “The hell with you.” (I didn’t say it, of course.) I met with the vice president of SDC shortly thereafter and said, “This is non-negotiable; I’m going to leave the company. I don’t want to discuss it, I don’t want to explain why. I’m just going to do it. I will stay as long as you like to hand it over.”

I stayed about 2 months. It was just too much¾too unrewarding to continue.

Ardito: You also said the work got boring, on top of this frustration. We know that Roger Summit was fascinated when adding features to the software...features such as the select and super-select commands, remove duplicates, and DIALINDEX, in which users could scan several databases to see where the information was. It wasn’t just loading databases.

Cuadra: Yes, there was continued development at SDC as well. For example, we added left-hand truncation, a document-ordering system, proximity searching, etc. I think we didn’t do proximity initially because we came out of the structured database context — NLM and MeSH — where searching free text wasn’t popular. The features that SDC introduced were a function of the context from which we came. We constantly added features. We would rudely proclaim new Dialog features, saying “another Dialog second.”

But the management situation engendered too much pain. There was no way to win. I had asked to hire outside computer services, but management wouldn’t let me. I wasn’t really serious — I just wanted some leverage, because there was no way to make money by taking in each other’s wash. The only way we could make money was by providing database access and attracting traffic. While the computer center could make money off of us, the company wasn’t making any money that way. It was handing it from one pocket to the other.

The Next Phase: Cuadra Associates

Bjørner: You didn’t give yourself much time to establish a consulting business when you left SDC.

Cuadra: No.

Bjørner: Did you make the decision to leave on your own and then later invite Judy Wanger?

Cuadra: I don’t think I told her in advance. I didn’t make the decision by myself; I made it with my family. I remember sitting at the dinner table and telling my wife Gloria and my two sons, who were 24 and 21, respectively. I said, “I’m thinking of leaving SDC; I don’t know what I’m going to do next. I’m sure I can get a job, but I don’t know what it is right now. Does that worry you?” They said “No,” and that’s all I needed. After I made the decision, I told Judy. I don’t know when Judy decided that she wasn’t going to work for the successor that was coming in. I think she left SDC the same day I did. I rented an office for Cuadra Associates and painted it the weekend after I left. Judy took a week off and came in the following week.

Bjørner: Did you have any clients before you left?

Cuadra: Not that I recall. We got one very shortly after, though — Haines Gaffner, of Link Resources — an entrepreneur if I ever saw one. He had sold the idea of a multiclient study on the online database industry, but he had no staff member who had any experience in this work. Somehow he found us and asked if we would help. At the time, we were the only people who had ever run an online service who were loose. We did much of the research and wrote the report, but Haines knew a lot that we didn’t know in other areas, so it was a gang effort. As part of the delivery, we gave a number of briefings, in New York and London, and there may have been others, but those are the two I remember.

Ardito: This was not a proprietary study?

Cuadra: Oh, yes, it was proprietary; the briefings were to the people who had paid Link. Haines and the other team members presented, so Judy and I were pretty visible. We had both been visible in SDC Search Service, so they knew that we knew something about online databases. During the first two or 3 years after we left SDC, we worked for probably 50 different organizations that needed some kind of consulting services. Some were database producers that wanted to start an online service; some were online services that wanted to be improved. We did some work for the European Economic Community, helping them identify ways to make them more competitive with their U.S. counterparts.

When we decided to build our product, STAR, and make it available to the market, we terminated our consulting business on the basis that we can’t be consultants if we’re selling a product. It would be conflict of interest, or the appearance of conflict.

Cuadra Directories

Ardito: Did the Link Resources study lead to the Cuadra Directory of Online Databases?

Cuadra: Yes. We learned of services that we had not known about before — nonbibliographic database services, such as GE Timesharing. We didn’t do any systematic analysis of those at the time, but we thought there were a lot and there was no easy way to find out what they were. So we decided to identify them systematically, and in order to make money, we gave seminars on nonbibliographic databases. Judy and another person, Ruth Landau, now Ruth Cuadra (she married my son Neil), traipsed around the U.S., Canada, and Europe, giving a seminar, with demos, called “Introduction to Nonbibliographic Database Services,” primarily to librarians. We invented the nomenclature. I remember Judy and I, and probably Ruth, fussing over the terminology and arguing….

Ardito: This has been a really useful classification that has stood the test of time.

Cuadra: The introduction of CD-ROM made us revise our terminology. After we had been publishing the Directory of Online Databases for awhile, the growth of CD-ROM and other portable media required that we develop another directory, the Directory of Portable Databases.

Bjørner: The terminology you chose to describe those is significant, because there are such different formats.

Cuadra: That’s right. At the time, there were various listings of CD-ROMs, of diskettes, of tapes, and of other media. We chose to take it up one level.

Ardito: So the term “portable,” in relation to databases, came from you.

Cuadra: Yes. The first portable databases Directory came out in 1990. I have it on my shelf. I also have a copy of the first Directory of Online Databases. It’s very skinny, with large type and wide margins. I still have the last one we produced, which has very small type and very small gutters.

Ardito: Can you tell us about the Elsevier relationship? Wasn’t there a Cuadra/Elsevier...?

Cuadra: We first started the Directory on our own; then we sold a half-interest to Elsevier. We published it jointly for a number of years, and then Elsevier decided it didn’t want to be selling to the library market.... It wasn’t paying off. [Laughter.]

The first issue of the Directory was done in hot type. We had no idea that we had a tiger by the tail until it was time for the upgrade — there were so many new online services, databases, and relationships. It was clear that we could never stay in business unless we could automate the process. We were building STAR for the library market, and we thought, “These are bibliographic records; we can use our own product to get a handle on this.” STAR became the machinery for putting the Directories out. The process got very, very efficient after awhile, and this got STAR into the publishing world. Without knowing it, we were building something that publishers needed to maintain their databases and spew them out electronically or in print, or later on CD-ROM. When Jim Monaco started Baseline in 1983, he ordered a copy of STAR; he was one of the very first users.

Competition

Ardito: Is your reaction to BRS during the early years as strong as Roger Summit’s?

Cuadra: I think I took less umbrage at things in general than Roger did. I just viewed BRS as a competitor. I believe they implied that they cared about their users more than others did. They had a committee or panel, but we also had regular users meetings; but we didn’t turn that into a marketing gimmick and imply that we cared more than others. BRS was just another competitor.

Bjørner: You didn’t have to deal with BRS for very long, actually. They came in toward the end of your tenure. And CD-ROM databases were not a threat to ORBIT while you were there. Do you have thoughts about CD-ROM as a medium?

Cuadra: I think of CD-ROM as kind of an electronic throwback to the time when one subscribed to journals. You had a certain budget and you subscribed to 23 journals. Then Roger and I and the others came along and found a way to find information in 10,000 journals without having to subscribe. The concept of CD-ROM is the same as a journal. You buy Psych Abstracts on CD-ROM, you buy MEDLINE on CD-ROM, you buy something else on CD-ROM. They have different formats; they have different names and fields. It doesn’t sound like a colossal event.

For organizations that use a single important database such as MEDLINE, using CD-ROM, which doesn’t require good telecommunications and expensive connect time, is useful. On the other hand, organizations that need access to 20, 30, 50, 100 different databases, can’t afford all of those CD-ROMs. Online access lets them get what they need, when they need it, and without a subscription; it’s a plus. Logically, CD-ROM is like subscribing to a serial. If all you need is that serial or two or three others, or if you’re overseas and have lousy telecommunications, CD-ROM is a great idea.

[At this point, longtime Cuadra associate Judith Wanger joined the conversation.]

Bjørner: Judy, we’ve been talking about when you left SDC and joined Carlos in Cuadra Associates.

Wanger: I think in yesterday’s world, we all thought we were part of a team that was going to go on forever. It wasn’t understood or appreciated that we were part of the future. ORBIT was an entrepreneurial enterprise within a government contracts-oriented company; the entrepreneurial factions were an anomaly. The business literature at the time was saying you’ve got to take these big companies and create little enclaves of entrepreneurs in order to keep your companies alive and well, but that was all theoretical. Now the companies actually do it. We were just too strange for SDC.

Ardito: Do you two regret leaving the online part of the industry?

Cuadra: No, I totally left the online service. My head stayed in it for awhile because of our directory publications — we were looking at it as if from 2,000 miles up. But we gave that up because it’s very hard to be a publisher of one title. The marketing staff it takes to sell one thing is the same staff it would take to sell 20 things.

Bjørner: You didn’t start any other publications? You didn’t want to be publishers?

Wanger: We did have a decision point with CD-ROM possibilities, the original optical technology. One of the options was to take the company in that direction. We assessed the opportunity and said, “This really isn’t what we primarily want to do.” We were creating STAR, and that posed a conflict. We shared some goals and some objectives, but publishing and software development were two different animals.

On to the Web

Wanger: The other opportunity we saw was that some of our customers looked at STAR when we first introduced the Web interface and said STAR could be a solution for making their collections available through the Web. Specifically, publishers and special libraries do that. So that was another possible avenue, because if you develop software for that, it is its own kind of software. You’re really going to focus on Web-based search engines. That would have taken us down a similar path as at SDC, and we didn’t want to go after that market.

Cuadra: Not a priority.

Wanger: Here’s something that is kind of illustrative of things that are happening with the Web today. One of our leading users early on was Mary Berger. She had a vice president who would come down to her — this was in a corporate library setting — and in corporate special libraries, there was already an orientation towards service. The librarians were experts in the field of their organization. They were marvels at finding things, but online became another tool in their arsenal. So this vice president asked a question. And had he not been there personally, as Mary likes to tell the story, she would have walked over to the shelf, pulled out a book, gotten an answer, and called him back. She would never have gone to the terminal and dialed in. But when he came, she would use online.

Ardito: The two of you have been philosophically in sync for such a long time. It’s quite extraordinary, considering all the things that this company could do, that you are in such agreement. What’s going to happen over the next 5 years? Could the Internet or its related technology change the direction of Cuadra Associates?

Cuadra: It’s intranets we think about now, as Judy indicates — the back room, the organization itself — not the hunt for data out there in the wide world.

You just reminded me of a story that I wanted to mention. Before there was communications security — in the olden days — SDC used to run some private files for individual customers that wanted good retrieval, but didn’t have a suitable program or a staff to build one. So did Dialog and so did BRS — private files, externally funded. Customers would contract with SDC; we would load their data and call this a private file. Only two people would know the password: the owner of the data and me (and maybe the programmer). We thought that was reasonably secure.

I was in London in 1976 or 1977 at the home of Nancy Vaupel, who used to work for SDC Search Service. At that time, we could dial into our computer and get messages. So, I dialed into our computer to get my messages, and out came some strange stuff. It was a report of the traffic for the past month on the New York Times Information Bank. At first I didn’t know what I was looking at. It just came out on my Texas Instruments 720 terminal¾click, click, click. “Why are they telling me this stuff?” I wondered, “I just want my mail.” I read it, of course, even though a gentleman doesn’t read another gentleman’s mail. I read it because I didn’t know at first what it was, and then I realized somehow I had gotten into the New York Times system. I didn’t have a password, I had no idea how to search it, I knew nothing about it, but it gave me all their most private stuff.

I realized that day that this is a very dangerous and insecure business, and that’s the way I feel about the Internet today. I don’t put a credit card on it. I’ve signed up for a lot of things, filling in blanks, and the format is so primitive. I’ve run into a dozen interactions that are mindlessly presented.

Bjørner: Do you think that traditional online is passé in the Internet environment? Do you think people are losing interest, that they won’t pay for the knowledge, the search?

Cuadra: I think about what I’ve seen with one of our customers. First they had five libraries; then they were down to one library, then no library. The departure of the personnel meant the departure of the skilled searchers. I don’t know why it happened — it may be that management thinks all the information they need is now free, and it will cost less if the end users themselves do all the searching — which is a dumb idea. With Chemical Abstracts, for example, there’s just no way anyone’s going to take on the job of identifying and naming all the chemical structures and do all that for free. The use of skilled searchers will decline to the extent that comparable data are available free, and it will decline to the extent that the people who know how to search are fired, and the libraries are closed.

When I first started teaching online searching classes at UCLA, I used something like 40 hours per student of connect time for the course. Up until then, the average had been only 3 hours. [Laughter.]

I’m going to tell you one more thing, which I learned from Judy. She gave a talk in England to pharmaceutical folks, then went to France, and ended up in Italy. To her shock and amazement, she discovered that one of the online services was looking at the searches of their users. At SDC, we had many customers who cared about patents and the priority of discoveries, and the idea that anyone would look at even their subject terms was just anathema. Judy saw that staff member of this online service were sitting at the console looking at what their customers were doing, and we realized that the capability was built into the software. I don’t know if the capabilities were ever used for purposes other than to see how the resources were being used, but this was a major ethical issue to us. I even remember when I learned about this, saying at some professional meeting, without attributing it, “We do not spy on our users. We have no software, and we will never build any that watches what any users type or what they’re searching for.” It was a major philosophical difference between us and others.

Wanger: You realize nothing’s new under the sun. Here we are addressing this issue of privacy, and it’s a hot one on the Internet today.

Cuadra: There’s something happening in the Internet world that I do think is important. It’s moving away from just finding information to integrating the content with some kind of analysis and tools. The pioneers for this were folks like Business International Corporation, which supported econometric modeling. You find a time series, but you perform analysis on it and compare it to some other time series and actually get results. The things they are building at Chemical Abstracts Service now involve the same concept, where you start with a reference and link to a full-text document. If you say “I want to find out more about the compound mentioned here,” it points you to the registry number that you can use to find more information and do various kinds of counting, sorting, and slicing. It’s no longer just, “Here’s the information; here’s an abstract.” There is a tool to help you do your work. I don’t expect the people who build the Yahoo!s and crawlers to understand how to do that, because to do that you have to understand the target audience and what they need to do and what their jobs are. So it’s very different from just finding cheap content and helping yourself.

Historical Records

Ardito: To conclude our interview, have you decided what you are going to do with your papers?

Cuadra: I have hardly any papers or artifacts from SDC. Essentially, I left everything there.
I have just a few miscellaneous things. I’ve been looking through them and found a few items that pertain to our discussion.

In 1956, I wrote a paper I’m very proud of: “Sources of Ambiguity in Psychological Reports.” This was when I was a psychologist at the Veteran’s Administration Hospital in Downey, Illinois. The psychologists there would prepare diagnostic, test, and other reports. I was curious as to whether our messages were getting through, so I designed an experiment. I took existing reports — for example, a report might say, “This person’s dangerous.” Then I would make up multiple-choice questions asking whether the report said that, a) this person ought to be president, b) this person is a danger to others, or c) this person is dangerous only to himself. I gave the test and the report to psychiatrists, student nurses, and psychologists with the directions, “Read the report and check off what it actually said.” I had the person who wrote the report do the same thing, so I had a criterion of what the writer was trying to say. The responses showed that only about half of the messages were getting through to the psychiatrists, social workers, nurses, and psychologists for whom the reports were intended. The reason I’m particularly proud of that paper is because my experiment addressed a major problem and was the first time that psychological reports were reviewed systematically. The funny thing is, my paper was accepted for presentation at the American Psychological Association (APA), but then APA changed its mind and “unaccepted” it because it would be too damaging to the profession. They didn’t want it on the program. But APA couldn’t control the printed publication, and the paper was published in the Journal of Clinical Psychology.

While I was still at SDC, we did a study of the impact of online services. This was part of our promotion efforts in 1974-75. We were trying to find out what good online access was doing. We got 10 organizations who either gave us their customer list or mailed surveys to their customers for us. It was the first time that anyone did empirical research on the impact of using online services. In addition to SDC and Lockheed, there was AEC, Battelle, the DDC, ESA, the National Science Library of Canada, NLM, NASA, and SUNY (it must have been before BRS). One of the things the study showed was that 25 percent of the libraries were offering online to provide reference service that they had never done o before. They would actually search for information they didn’t have in the library. That was one of the real surprises: They were able to provide a type of service that had never been done before.

Judy Wanger did the largest part of this research; her name is first on the study. There was a chapter on expectations versus real gains — with a chart on the number of end users being served — and it shows the number increasing. There was a warning that anyone deciding to start online service should be prepared for the possibility that the demand would go up. This was widely quoted in the library world. There are at least two reviews of it, one in Library Quarterly, and there was one in Japanese.

Challenging Early Perceptions

I’ve discovered that I was rather pugnacious about online. I talked regularly on the growth of online databases and, initially, on online access. I found one paper that I gave in 1977, at a Pittsburgh conference, where Allen Kent referred to online in terms of a “bandwagon.” As I looked through my paper, I realized that I really took him apart on this and said, in effect, that to use online is a decision being made by intelligent librarians who realize that it will improve services to their patrons. I was surprised at the language I used in responding to comments by a venerable figure.

Our study found that the introduction of online service increased the prestige and respect given to library professionals. I used that result to comment on Allen’s paper: “I feel a need to comment on the idea in Allen’s paper that buying and selling information service constitutes prostitution. I find both the concept and the language in the way it’s expressed (by others, not Allen) as uninformed, confused and unprofessional — I’m afraid that a simplistic and rigid approach to the idea of equal access can do a disservice to the library profession. As for libraries that curtail the introduction of sophisticated new services or tools until their budgets are large enough to offer them free to everyone, I foresee a continued erosion of the library’s position in the total information environment in the United States.”

So I was quite pugnacious¾I was much braver then.

In the next segment, we continue this series with an in-depth interview with Roger Summit of Dialog. The authors would appreciate hearing readers’ experiences of their early days with online searching and their views of the development of the industry. Please send e-mail to the authors directly (see below) or to Barbara Quint at Searcher magazine [bquint@mindspring.com].