On the other hand, a national information system can be a means of tying together the widely separated and separately administered programs of the many governmental and private groups. By vesting responsibility for a national information system in a given area of science and technology in a single Federal department or agency, a long step forward will be taken toward more effective coordination of the Government-wide research and development effort.

I believe that the effective coordination and marshaling of Federal activities in a given information subject area generally will not require significant revision of present departmental statutory bases so much as it will require some special supplementary legislation_which will reflect the intent of the Congress with respect to extra-departmental issues and opportunities. This is especially important with national information systems because they represent a coalition of both public and private activity and have a significant impact on all sectors. The problem of national information systems is to recognize the motivation of mission-oriented departments and yet provide an effective managerial and fiscal mechanism by which measurable progress toward a national objective can be achieved.

We have much to learn about the building and operation of national information systems. My remarks have attempted to communicate to you what we have learned to date in this area and my conviction that legislation such as the proposed bill, H.R. 15638, is necessary, to articulate activities and provide focal points of responsibility and attention on national goals.

I believe that the realization of national information systems is one of the most challenging areas of current activity, with great potential impact on not only our science and technology but also on the academic, commercial, and industrial sectors of our society. I believe the Standard Reference Data System is an important beginning and that the enactment of H.R. 15638 will make it possible to realize its potential. I will be pleased to answer any question, Mr. Chairman. Thank you.

Mr. DADDARIO. Thank you, Dr. Hornig.

I have a question that involves the idea that you express at the bottom of page 7. You say:

I consider it to be very important to encourage and support the information handling activities in the private sector.

Could you give this committee an idea as to what you mean, and how you think it might develop?

Dr. HORNIG. Well, for instance, and I am now speaking of informa tion in general, rather than just the Standard Reference Data System, the chemical literature is published by the American Chemical Society. I don't know the exact count, but they publish some 15 or so different journals in various areas of chemistry.

In physics there is the American Physical Society that publishes the original literature. The American Biological Society publishes biological literature. The Engineering Society publishes engineering journals. So these are all private activities. In fact, the primary scientific literature publication is entirely in private hands. But they have great difficulties. There are many problems of coordination.

The most difficult one is I mean information that is published in the chemical literature-how does it get into the hands of an engineer

who doesn't read the very specialized publications of the Chemical Society, for example. This is what produces the problem.

So we have very many journals, but most practicing engineers and very few scientists, even if they read 20 or 30 journals a month, can even dent the total amount of literature publication.

Then there are many inventions to deal with this. The Chemical Society, for example, publishes something called Chemical Abstracts. The Physical Society publishes Physics Abstracts. These are attempts to publish small abstracts of the journals and classify them by author, subject, and so on, so people can find them. Well, the number of abstracts gets so excessive that they then publish annual indexes to the abstracts.

They used to publish 10-year indexes to the indexes but now the volume of data has gotten so great that the Chemical Society has simply decided it can't publish any more decennial indexes. It is just overwhelmed. It can't afford it.

The question is they are giving up. So we are working together. There is now a program between the Science Foundation, the National Institutes of Health, the Department of the Army, and Chemical Abstracts to work on means of developing computerized ways of handling this information. So this is an example of Federal-private cooperation. This indexing is absolutely essential because otherwise the people who need it, who are the engineers, the people in industry, the people in universities, won't have access to the work we have paid quite a lot of money for.

Mr. DADDARIO. You see nothing in this bill which would bar the private sector from participating in this activity?

Dr. HORNIG. Oh, no. There are, of course, excellent examples in the past of the cooperation of the Bureau of Standards with the American Petroleum Institute in putting together critical data (thermodynamic data), the data on hydrocarbons. It was the Petroleum Institute which collected the standard samples of hydrocarbons from which the data were taken that made the compilation. So there are many examples of cooperation.

Also in the sense that Mr. Vivian mentioned, where much of the data originates now in the first place in private laboratories, but which don't themselves compile data. So someone has to do something with it.

Of course, some of it they withhold for proprietary reasons, but much of it they are perfectly willing to put into standard manuals and cooperate, if there is only someone to do it.

Mr. DADDARIO. Dr. Hornig, when this committee first met and briefed itself on the aspects of this legislation, a question came up as to whether or not by compiling and making this information available, that the Government was also guaranteeing it. What are your thoughts on this?

Dr. HORNIG. I don't think so, except in a certain sense, as with a private activity, like when the International Critical Tables were put together, the authors said that this represents their best estimate of what is most reliable. They cite alternative values, or alternative

sources.

I think this is the same responsibility the Bureau takes when it publishes something like this, with or without a mark. It says: “In our

judgment and according to our best people this is the most reliable thing you can find.”

Now I have found myself in using such data in the past, that I don't always agree with the evaluator. Sometimes if it matters particularly to you, you can take the effort on a single number to go back and check yourself and see whether you have an independent judgment. So, of course, it takes a responsibility. This is what you mean by critically evaluated data. Someone has to be responsible for the critical evaluation and whoever was responsible has assumed a very serious responsibility.

But I don't think that the provision, for example, of a mark changes that responsibility, nor does it constitute a guarantee in any sense. Every scientist and engineer knows that many of the numbers he considered thoroughly reliable 10 years ago have changed. I don't think he is likely to be deceived.

Mr. DADDARIO. Dr. Hornig, on page 2 of your statement you say:

In many cases, these compilations find their way into the general public use, but in many other cases, they remain of proprietary interest.

Do you mean that they remain as a "proprietary interest" unintentionally?

Dr. HORNIG. Oh, no; not accidentally. I mean if I am in an industrial laboratory and I have developed a new alloy, for example, which I am using in some special way in my processes and have spent quite a lot of money characterizing it very precisely then I may consider it to be quite important not to disseminate it while it is of commercial value to me.

Mr. DADDARIO. It would seem to me that if this is a matter of proprietary interest, you would see to it that it would not get into a national system. Can you see anything particularly wrong with that?

Dr. HORNIG. That is right. But besides the ones that are withheld as a matter of proprietary interest there are many more which might, if there is a receiver, someone to talk to them, very well find their way into the public domain.

Mr. DADDARIO. Then you really mean that as the private sector develops information in this area, there is certain data which they would make available and there is other data of a proprietary interest which would not be made available.

Dr. HORNIG. That is correct. And there is also the intermediate case where one can get it, where they maintain it in proprietary form but are willing to give critical judgments on the basis of their experience as to the value of the data which you plan to publish. If I am in a private laboratory and I know the published data is wrong, I may tell you that, even if I am not willing to tell you what is right. And this may be very important to do just that much.

Mr. CONABLE. Mr. Chairman.

Mr. DADDARIO. Mr. Conable.

Mr. CONABLE. Following up this line of questioning, Dr. Hornig, I ent copies of this bill to some scientists in my area. One man who is a research scientist for General Dynamics, Mr. R. A. Santirocco, has written me a letter. I would like to read you a paragraph from it and ask your comment.

He says:

Section 7(b), restricting the "copying" of any Standard Reference Data compilation, needs clarification. The language of this section appears to contradict the well-established rule that Government publications cannot be copyrighted. The scientific community will, I assure you, be highly antipathetic to any statutory restrictions on free dissemination of the data.

Even more outrageous is Section 5, providing that Standard Reference Data "may be sold" by the Secretary of Commerce. Nothing could be more clearly in the public domain than scientific data, collected in most cases from the open scientific literature at public expense, by a non-Defense agency of the Federal Government. To the extent that any rights can exist in facts, these rights plainly reside in the people of the United States. The notion that the Secretary of Commerce can sell them is absurd. It must be deleted from any final version of the bill.

Will you comment on these remarks, sir?

(The complete letter follows:)

Hon. BARBER B. CONABLE, Jr.,

House of Representatives,

Washington, D.C.

GENERAL DYNAMICS ELECTRONICS,

Rochester, June 21, 1966.

DEAR BARBER: I have read H.R. 15638 and the attached Commerce Department analysis with great interest.

The intent of the bill is probably good.

Commerce has not overstated its case on the scientist's need for compilations of reliable reference data. I am not sure, however, that this bill will remedy the situation.

In particular, I question the "critical evaluation" function of the proposed Standard Reference Data project. The bill would set up elite panels, paid fat consulting fees, to evaluate the relative accuracy of the various data culled from the scientific literature. The processes of science, however, demand that the individual researcher make his own judgments, not that he rely on the packaged, "certified" opinions of anonymous, bureaucratically controlled expert panels. Science and the public will get the most for their money if the Standard Reference Data service concentrates solely on the production, and yearly updating, of comprehensive annotated bibliographies concerning the data in question.

Incidentally, there is a semantic pitfall in the word "standard." This word suggests something conventional, arrived at by mutual agreement. But the data we are considering are objective facts. Perhaps the word "standard" should be deleted everywhere in the bill.

I am shocked by the legal defects in the bill. Section 7(b), restricting the "copying" of any Standard Reference Data compilation, needs clarification. The language of this section appears to contradict the well-established rule that Government publications cannot be copyrighted. The scientific community will, I assure you, be highly antipathetic to any statutory restrictions on free dissemination of the data.

Even more outrageous is Section 5, providing that Standard Reference Data "may be sold" by the Secretary of Commerce. Nothing could be more clearly in the public domain than scientific data, collected in most cases from the open scientific literature at public expense, by a non-Defense agency of the Federal Government. To the extent that any rights can exist in facts, these rights plainly reside in the people of the United States. The notion that the Secretary of Commerce can sell them is absurd. It must be deleted from any final version of the bill.

In summary, I believe that scientists and engineers will be sympathetic to the aims of the bill, but that science and the public will benefit most if the bibliographic, rather than the critical, function is made paramount. However, few scientists, engineers, industries, or universities will support the bill with Sections 5 and 7(b) as they now stand.

I appreciate your thoughtfulness in soliciting my opinion, and I hope this has been of some service.

Sincerely,

R. A. SANTIROCCO,

Research Specialist, Special Programs Laboratory.

65-891-66- -2

Dr. HORNIG. This question of user charges is a very difficult one. I think when Dr. Hollomon is here you should discuss with him the more detailed thoughts of the Bureau of Standards. I think the idea of charging for a number, if that were the question, would be absurd. But a good part of the expense resides in the effort to collect, compile, and set up special filing-I mean computer services and so on. It is my view that the use of reasonable user charges is one of the best ways of determining the value of services performed and that user charges as a concept does make sense under many circumstances.

There are the general compilations which one wants as widely disseminated as possible but there will be cases where special compilations in special areas have particularly value to particular groups. I think in such cases that user charges to me do make sense.

I think what I am saying is that if the bill said that there would be user charges for all compilations, this might be a justifiable criticism, but the bill doesn't say that. The bill allows user charges in appropriate circumstances.

Mr. CONABLE. This really is establishing a proprietary interest in the Government when any sort of user charge is allowed, is it not?

Dr. HORNIG. In a certain sense that is true. But you know when you establish computer services, for instance, one can set up enormously expensive systems for dissemination of information, and one of the problems is deciding what is really useful to people, what services are really useful, and which ones are nice, just desirable, and when they get expensive one of the best ways of finding out what people really want is what they are willing to pay for.

I think it then becomes a matter of more detailed policy, whether one charges full costs or charges enough so that one doesn't perform useless services so it has to be of some real value to the user if you are to perform this service.

Mr. CONABLE. That is all.

Mr. DADDARIO. We are happy to have the chairman of the full committee here. This is his bill, and he may have some questions?

Mr. MILLER. I have no questions, Mr. Chairman. I am very happy to see Dr. Hornig, and also Dr. Astin who is in the audience.

Mr. DADDARIO. We are happy to have them here too, Mr. Chairman. Mr. Roush.

Mr. ROUSH. No questions, Mr. Chairman.

Mr. DADDARIO. Mr. Mosher.

Mr. MOSHER. Mr. Chairman, you earlier raised the question of what the costs would be and what the returns might be. Dr. Hornig indicated that this is very difficult to determine and that the returns might be 10 to 1 or 100 to 1. If I interpreted you correctly, Dr. Hornig, you indicated that regardless of what the returns may be, in your mind, this is an essential bill-almost an imperative thing to do. Dr. HORNIG. I am not sure that the two are different. I do think it is essential, but I think it is essential because the returns are great. Mr. MOSHER. I agree.

Dr. HORNIG. I don't think we could proceed with this enterprise without this data.

Mr. MOSHER. In other words, it isn't necessary to compute these returns with greater accuracy. In your mind they are so obviously