can Bar Association and the Association of the Bar of the City of New York be accepted into the record since it compares the present copyright law with the proposed statute as they affect computer usage of copyrighted material. Information storage and retrieval began as a response to the flood of data and reports emerging from the enormously expanded research efforts of the past two decades. Currently, it is being applied not only to research but to education as well. Its basic instrument is the computer which stores information on magnetic tape, but information systems are based also on microfilm and punchcards of several kinds.

At Bethesda, Md., stands the National Library of Medicine. It houses a sophisticated computer system and a highly trained staff who reads and indexes all of the articles appearing in current medical journals and books. The computer collates and prints out the Index Medicus, which is published and distributed on a regular schedule. This information system, known as "Medlars", is primarily devoted to indexing.

A related but significantly different information retrieval system was announced in the New York Times on March 4, 1965. I quote:

"The medical libraries of three major eastern universities will be tied together in a network of computers and telephone lines to give scholars virtually instant access to their pooled resources * * *

"The libraries, at Harvard, Columbia, and Yale Universities, are ranked among the largest and best in the world. They are in the primary stages of electronic integration--and five other libraries have been invited to join the system.

"Although there is much duplication, the three libraries will then contain 1,025,000 items. These can be searched by computers in seconds *

"The system under development here, with the help of a grant by the National Science Foundation will be equally adaptable to groups of law, business, or general libraries ***

"When telecommunication and photographic reproducing devices are added to the network system, it will be possible to eliminate some duplication of material among libraries. Pages from a book in New York could be flashed to a user in another city and even reproduced for him in take-home form."

Later in the same news item, Thomas P. Fleming, head of the Columbia University Medical Library, states:

"With the computer you can enter many, many more subject headings. These go in compact, miniaturized form on magnetic tape and are stored on disks. Just as in a jukebox when a person pushes the button, the arm finds what he wants, so the searching arm goes along with the disks and picks out what the researcher wants." [Emphasis added to call attention to what is clearly analogous-the jukebox exemption of the current law.]

The difference between "Medlars" and the medical library network is clear. At present, "Medlars" is devoted primarily to the indexing of journal articles. The new network is designed to retrieve and print out information from books. "Current Research and Development in Scientific Documentation," volume No. 13, dated November 1964, issued by the National Science Foundation updates similar volumes previously published. In close to 500 pages, the current status of 150 information storage and retrieval systems is described. These include the activities not only of Government-sponsored research projects of universities, but of many of our major corporations such as Radio Corp. of America, Philco Corp., Motorola, Sperry Rand, IBM, ITEK, Bell Telephone Laboratories, Burroughs, Magnavox, and Chrysler Corp.

May I call your attention to page 243 of this publication which reports on the Recordak Corp., a subsidiary of the Eastman Kodak Co. The Recordak Lode Star Reader-Printer described in this report is in operation here in Washington. It makes possible automatic location and reproduction either on a screen or on paper the pages of a book or journal located in a central storage miles away. I have seen this system in operation. The access time for locating a document out of many thousands and flashing it on the screen is approximately 7 seconds. Print-out appeared to me to be as rapid.

It should be noted that photoduplication, or print-out, as it is called, is part of the service rendered by information retrieval systems. At previous sessions of this committee you have heard requests for free (unpaid for) photocopying of copyrighted materials. I respectfully urge that you relate these requests to the mechanisms and practices of information retrieval.

How do information retrieval systems use copyrighted materials? The new science of documentation has already produced a highly technical vocabulary,

but the meanings of the words are fairly clear. The steps in the process of information handling are as follows:

Input: Text and data are fed into a computer. The input may be both indexes and documents. The documents may be tables of data, journal articles, whole books, or graphic material such as maps, charts, and infrared spectra.

Storage: The material fed into the machine is recorded on microfilm or tape at fixed "addresses."

Retrieval: By reference to an index of these locations, or addresses, the machine scans and searches the content of its memory, locating the information required.

Output: In response to request for specific information, the stored index entries, bibliography, or the documents themselves may be flashed on a screen or printed out as hard copy. Appropriate excerpts, paragraphs, or whole pages are made available either in microfilm or in normal-sized printed form. In short, a copy is made.

Let us consider these steps in terms of copyright law. Some of the techniques of input involve the translation of the content of a copyrighted work onto punchcards by means of a highly specialized electric typewriter. The copyrighted work is then transferred from the cards into electrical impulses which are stored on a magnetic tape. Thus the copyrighted work can in whole or in part, in its original form or in a derivative form such as an abstract, digest, or condensation, be stored in a memory core of the computer.

Computer programing, for which copyright protection is urged, is the organization of data and copyrighted works and the creation of instructions so that these works would be available for retrieval. What the computer programer is really doing is creating a new kind of anthology or compilation of copyrighted works that is published in a computer system rather than in the conventional bound book form. The storing of the anthology, compilation, or selections in the computer is merely a newer and more sophisticated use of the publishing or copying right which clearly inures to the copyright owner.

There is a question as to whether the scanning of copyrighted works in the "memory core," for the purpose of selecting appropriate sections, is subject to the exclusive rights of the copyright proprietor. This question must receive serious consideration.

Output or transmitting the copyrighted work in response to an inquiry for information may take many forms. It may consist of flashing a copy of the selections on a screen or a duplicate of the excerpts or pages may be printed out. In the area of text or reference books, the response rarely requires more than an excerpt, a paragraph or some other relatively small portion of the whole work. All the user requires is this small portion flashed on the screen or printed out as photocopy in response to pressing a button on the “jukebox" as Dr. Fleming, of Columbia University, calls the computer system. Thus, computer output is another phase of information storage and retrieval publishing.

If one supposes that information storage and retrieval systems are not now and are not likely to become commonplace during our lifetime, then we may leave the resolution of the rights of a copyright proprietor in respect of computer usage as a legacy to future generations. But in doing so, would we not be creating another jukebox exemption, this time one of even greater consequences to the world of private authorship and private commercial publishing? The hard, inexorable facts are, that information storage and retrieval systems do exist and are increasing numerically as well as in the scope of their uses. The Wall Street Journal dated May 3, 1965, carried a lead article entitled "Selling to Schools: Big Educational Outlays, New Teaching Methods Create A Vast Market. IBM Develops Computerized Learning Systems: Steel, Carpet Firms Also Score."

I quote:

"As school enrollments swell, such big and technically advanced companies as International Business Machines Corp., General Telephone & Electronics Corp., and Minnesota Mining & Manufacturing Co. are expanding efforts to develop and market school products. For some, educational research offers a way to keep laboratories busy when defense contracts are hard to come by *

"Many of these companies have been attracted by a revolution in teaching techniques and their selling zeal and technical skill are now hastening this. The "knowledge explosion" and a shortage of qualified teachers are driving more and more schools to new instruction methods relying heavily on such electronic

tools as television, tape recorders, communications networks, and even computerized information storage."

The same Wall Street Journal article then continues:

"IBM scientists have developed an experienced computer 'language,' called Coursewriter, to enable teachers with no computer training to easily enter instructional materials into a computer. Then, students at typewriter-type keyboards anywhere from 10 feet to thousands of miles away can receive instruction and reply to it at any time they choose."

Three days later, in its May 6 issue, the Wall Street Journal reported the Westinghouse Electric Corp. statement that it has developed an electronic system that plays television pictures from a phonograph record and, according to the Westinghouse statement, "Pictures on the television screen may be line drawings, charts, printed text, or photographs. Applications include classroom instruction, industrial and vocational training, sales presentations, and remedial teaching."

I would like to cite for your consideration a National Education Association publication issued by the Department of Audiovisual Instruction entitled "The Role of the Computer in Future Instructional Systems."

The chapter contents of this NEA monograph are:

Computer fundamentals.

The computer-based teaching machine.

Information retrieval systems and education.

Computer-based simulation: a new technology for education.

The automated classroom.

What is next in computer applications to instruction?

The direction of events is clear. The issue reduces itself to one question: do we meet the technological revolution by conscripting copyrighted works? Now, let us turn to the less esoteric devices which I have included as part of the technological revolution-as affecting the economic core of authorship and publishing-I now speak of photocopying devices.

The development of photocopying devices, both in the numbers and kinds of machines and techniques available, as well as in the rapid reduction of costs of photocopying is equally startling and spectacular. Mr. C. Peter McColough, executive vice president of operations for the Xerox Corp., reported to the Los Angeles Society of Financial Analysts on Thusrday, June 3, 1965, as follows:

"As to the total copying market, its growth also continues at a rapid pace. In 1964, roughly nine and a half billion copies were produced in this country, resulting in total income for the industry of about $500 million. By 1969, I'd guess that some 25 billion impressions will be made by copiers. And the 'information explosion' will still be accelerating."

Mr. William R. Hawken, an author and authority in the field of photoduplication who for many years had been associated with the Eastman Kodak Co., wrote an article in 1954 entitled "New Methods for Photocopying” in which he described 12 machines. In the March 1963 issue of the magazine Administrative Management, W. A. Kleinschrod surveyed 174 machines.

In a letter dated June 16, 1965, addressed to me by Mr. Stanley Salmen, coordinator of university planning of Columbia University of the city of New York, he stated that in their library system which is used almost entirely for research they have bought this year 1,700,000 sheets of Xerox paper as compared with 837,000 the year before and 285,000 the year before that. Of course, not all of this photocopying was of copyrighted works. However, the increase from 285,000 sheets for the year 1963 to 1,700,000 in 1965 is indeed indicative of its extraordinarily rapid growth in library use. These statistics lay to rest as outdated the conclusion of the report by George Fry & Associates on photocopying completed in 1962 and based on experience prior to that date.

The significant fact about all of the devices for information retrieval and photocopying is this: they are designed to replace the printed book. They will make 1 book do where 100 had been used before.

The basic purpose of the first book published-in fact, the first clay tabletwas to record information, store it, and make it available for later retrieval. Nothing has changed since except the technology. Although they are referred to as devices for information storage and retrieval, they are in fact devices for printing and distributing literary works created and first published by others. Photocopying equipment, information storage and retrieval really fulfill the functions of a new, speedy, accessible, and more efficient reprint publishing house.

The only difference between clay tablets, conventional publishing processes and computers and photocopying is one of efficiency and convenience. There is no philosophic or economic logic to exempting from licensing or fee the use of an author's work simply because one device is substituted for another.

If the rights of authorship intended by copyright legislation are to remain real and of practical economic worth, the impatient urge for unimpeded access to copyrighted material cannot, must not, include exemptions in the copyright law that result in the disappearance of private authorship and publishing by authorizing free use of the fruits of their labor.

I did not come before you only to urge philosophical and intellectual conclusions. I have been authorized by the American Textbook Publishers Institute to describe a blanket licensing system for copyrighted materials which would satisfy this year's needs as well as the future needs of educators and researchers for unimpeded access to copyright material.

ATPI proposes to set up a clearinghouse for the reference and instructional materials published by its member firms. Their materials will be available to information retrieval systems now found chiefly in government and industry but soon to be installed in networks of college libraries and in our public schools. The materials published by ATPI members will also be available for photocopying devices wherever located. The proposed clearinghouse would be open to all users of copyrighted instructional materials. Our immediate present concern is with teachers at all levels of instruction.

We recognize that the needs of teachers extend to many other kinds of copyrighted materials. We believe that the example of a clearinghouse co-sponsored by ATPI and professional groups will mobilize other proprietary interests to similar action.

We believe that the present statement should not set forth fixed, full, and exact details since these should be developed in discussion between interested parties. We therefore set forth at this time only the general purposes of photocopying licenses.

General description

The clearinghouse, as we see it, would have the following characteristics:

1. It would provide a blanket license for photocopying copyrighted educational material for an annual fee to any school system and nonprofit educational institutions.

2. The blanket license would permit copying without limitations as to the number of copies or the extent of the material copies, short of an entire book. 3. It would make available the entire list of publications, in print or out of print, except for consumable materials such as workbooks, laboratory manuals, and standardized test answer media of the member firms of ATPI. The effects of such an arrangement, we believe, would be as follows:

1. To relieve teachers of any anxiety or uncertainty as to whether they might legally or ethically copy or reproduce copyrighted materials. 2. To relieve educators and publishers of the necessity for correspondence as to permissions.

3. To obviate attempts to define such difficult terms as "excerpts," "quotations," or "reasonable number of copies."

Procedure

1. We propose the immediate formation of an advisory council to the clearinghouse to be composed equally of representatives from the professional community and from ATPI members. This would be a permanent committee with representatives chosen for terms long enough to make their work effective.

The functions of this council would include the following:

(a) Developing a plan for organization and operation of the clearinghouse;

(b) Receiving grievances from publishers and licensees; and
(c) Arbitrating between the clearinghouse and the licensees.

2. The licensing agency, or clearinghouse, would be operated by ATPI. The files, records, and books of account of the clearinghouse would be open to the advisory council at all times. The director of the clearinghouse would be required to submit detailed quarterly operating statements to the advisory council.

3. Since user fees would not be on a per-use basis, it is essential to establish a sampling structure to determine a fair base for the blanket fees. The need for sampling will be continuous. It is anticipated that all income derived in the first 5 years would be devoted to sampling studies.

Members of ATPI have already lent financial support to the Committee To Investigate Copyright Problems Affecting Communication in Science and Education's project to engage the Social Science Research Institute of Washington, D.C., in a sampling study. Dr. Howard A. Meyerhoff, president of this committee, scheduled to testify today, will undoubtedly inform you more fully of the status of this project.

It is with this kind of system in mind (or any other suitable one that might be suggested) resulting both in fair payment to authors and publishers and ready access to teachers and students that we make the following specific comments on the language of H.R. 4347.

May I call your attention to section 109 and restate the point of view expressed by Mr. Deighton in his testimony of May 26 as it pertains to information retrieval. Even though the Register of Copyrights Supplemental Report of May 1965 would suggest a more limited intent, the exemptions afforded in subsections (1) and (2) of section 109 could in fact permit educators to use all copyrighted materials in information retrieval systems without permission or payment to the copyright owners. This freedom is afforded by relating the applicable definitions in sections 101 and 106 to section 109. Section 106, subsection (b), at lines 32 through 38, reads as follows:

"(1) To 'perform' a work means to recite, render, play, dance, or act it, either directly or by means of any device or process or, in the case of a motion picture, to show its images or to make the sounds accompanying it audible."

"(2). To exhibit' a work means to show a copy of it, either directly or by means of motion picture films, slides, television images, or any other device or process." [Emphasis added.]

Section 101 at page 3, lines 3 and 4, informs us that:

"A 'device,' 'machine,' or 'process' is one now known or later developed."

Thus, all present and future uses of copyrighted material by information storage and retrieval mechanisms are covered by these definitions and apply to the proposed exemptions set forth in section 109.

Moreover, the definition "to transmit" set forth in section 101, at lines 13 through 15 reads: "To 'transmit' a performance or exhibition is to communicate it by any device or process whereby images or sounds are received beyond the place from which they are sent."

Since the definition of "to transmit" does not expressly refer to radio and television broadcasting, the language describes information storage and retrieval systems as well.

Thus, the language of these three definitions when coupled with the exemptions set forth in section 109, would permit the transmission of copyrighted material from a central storage location to any retrieval station located anywhere in the United States, or in the world for that matter, provided that, as section 109(2) states, it was primarily for reception "in classrooms or similar places normally devoted to instruction and is a regular part of the systematic instructional activities of a nonprofit educational institution."

The ad hoc committee's (of educational institutions and organizations on copyright law revision) proposed section 111 seeks to broaden even further these exemptions.

In explaining the effect of their proposed section 111, the ad hoc committee states:

"Where section 109 authorizes a teacher to perform, exhibit, or transmit a copyrighted work, the proposed section 111 would permit the teacher to make

"(2) a reasonable number of copies or phonorecords of excerpts or quotations from that work but only where the excerpts or quotations are 'not substantial in length in proportion to their source.'"

When added to the permission of performing, exhibiting, and transmitting the work by information storage and retrieval already exempted in section 109 the result would be that copyrighted works could be used in all phases of information storage and retrieval and the copyright owner completely bypassed if such uses were for educational purposes. This includes information storage and retrieval systems as we know them today; the classroom of tomorrow as the NEA monograph describes it; and the libraries of tomorrow as General Electric Co., National Science Foundation, the Council on Library Resources, Massachusetts Institute of Technology, insist are inevitably on the way.