sion bill makes it a copyright infringement to input copyrighted materials into a storage and retrieval system without regard to the possibility that the copyrighted materials may never be used in a way that will have any economic effect on the copyright proprietor. The mere act of taking a copyrighted work, putting it on punchcards or magnetic tape, and storing it in a data bank-without more-is an infringement under Senate bill 597. If permission is sought, presumably there will be a copyright toll on it.

It must be remembered that education buys books. It buys enormous quantities of them and in recent years has been buying them in everincreasing numbers. As a result, education pays the publishing industry enormous sums for its couyrighted products. Many of these books are put on library shelves. The fact that they are stored on library shelves and retrieved by faculty and students from library shelves has never been thought to justify the imposition of a second copyright payment. The fact that people go to library shelves and search for what they want, pull down or retrieve the books that strike their fancy, flip and scan pages, read entire books and think about their contents, and then restore the books on the library shelves does not bring down a second copyright toll. Yet, Senate bill 597 says that if we choose to store books in a data storage and retrieval system, as opposed to on library shelves, a second copyright tariff will be imposed. Senate bill 597 announces to teachers, scholars, and students, if you want to avoid being charged twice for gaining access to a work, you had better be prepared to turn your back on the new technology and go to the library shelves, lick your thumb, flip pages, and do your research in the old, manual, and primitive way. Don't ask the computer to help you; that will be a copyright infringement.

I submit, Mr. Chairman, that this does not make any sense. What right does the publishing industry have to extract a payment at input? Did they invent data processing technology? I think Senator Fong put his finger on it most penetratingly yesterday in a colloquy with Mrs. Linden. He suggested that charging for input without regard to the possibility that nothing of copyright or economic significance will ever be done with the stored material is like a restaurant imposing a cover charge for looking at the menu. Senate bill 597 effectively charges the computer user for the privilege of determining whether he wishes to eat any part of the copyright pie.

In summary, in two respects, the question of input and the question of the exemption in section 110(2) (D), the ad hoc committee thinks that the present copyright bill very badly restricts the development of computer technology and the use of this machinery in the most efficient and beneficial way possible.

It is perfectly clear that the publishers fear the capacities of data processing machinery. Indeed, we heard a great deal about their fears this morning. They fear the possibility that sometime in the future there will be a national computer network, tying together every major university in the country, so that with the press of the proverbial button, Dean Siebert's students in Michigan State University can dip into the data storage bank at Harvard University, or my class at the Michigan Law School can reach into a bar association library in New York to retrieve the full text of a book.

This is all possible, Senator. We do not deny it. But its realization is far in the future-far in the future. We are going to experience a long gestation period during which computer users will do little more than tinker and experiment on a very, very limited basis. The existing proposal, which is drafted in terms of the capacity of the distant future, will hamper the tinkering and experimentation that is so crucial in the immediate future.

Senator BURDICK. Is not the thesis of Mrs. Linden yesterday, that it is all far in the future and we might be acting prematurely at this time?

Mr. MILLER. That is correct, Senator. Might I suggest that the Congress would be acting prematurely by enacting the provisions in Senate bill 597. In other words, the crux of the matter seems to be a dispute over the structure of the Copyright Act pending a study in depth of the problems created by data processing and an analysis and development of a rational accommodation among education, the publishing industry, and the computer hardware companies. The ad hoc committee believes that pending some rational evaluation of these enormously complex and important questions, the Congress should maintain the traditional exemptions accorded to education. The fear of a fullblown university data network is an unreal one at this time.

Moreover, nothing should be done at this time to prejudice the development of the new technology. Educators and others should be encouraged to experiment and try to ascertain the capacity of the new data processing devices. Let them do so while the much needed study is in progress. That is the essence of the proposal laid before this subcommittee by the ad hoc committee with regard to the text of section 110. It is appended to Dr. Wigren's statement.

If I might shift my attention to the question of computer programs. Here again, the proposed legislation speaks and it speaks without the benefit of a legislative record. Several people appeared before the House committee and suggested that copyright protection be extended to computer programs. Their testimony was not challenged and the implications of the suggestion were not tested. Their position in effect has been adopted by Senate bill 597 because the definition of "copyrightable works" in section 102 includes any original work of authorship "fixed in any tangible medium of expression," which "can be perceived, produced, or otherwise communicated, either directly or with the aid of a machine or device." That means that if a computer program is put onto punchcards, magnetic tape, disks, or any other vehicle, it is copyrightable because it has been fixed in a tangible medium of expression and can be perceived by machines.

It is the position of the ad hoc committee, and I am quoting from a resolution adopted by the committee, "That a computer program— that is, the instruction from the computer as distinguished from the data stored in the computer-should not be subject to copyright."

The concept "computer program" covers any set of instructions or commands that are directed to a computer or data processing machine. A program can monitor the operations of a steel mill, for example. It can compile an inventory or payroll or prepare an index or concordance of someone's poetry. In short, programs can do a multitude of things. The important point to recognize is that a computer program

is a functional item. It is a mathematical and logical sequence of instructions or commands to a mechanical device and should be distinguished from books or plays or motion pictures or poetry-the forms of expression that traditionally have been covered by our copyright legislation.

The vice of granting a copyright to a computer program is the following: if you recognize a copyright in a program, it is quite conceivable-indeed, the broad language in sections 102 and 106 of the proposed legislation seem to so indicate that the copyright protection may extend to or embody the process, scheme, or plan that the program uses to achieve a functional goal. This means that if somebody develops a program for running a steel mill, the scheme or process captured on the program might become subject to copyright appropriation. This has never been considered a proper function of copyright law in this country. Indeed, the Supreme Court has indicated under an earlier copyright act that copyright protection does not go this far. In effect, a copyright that extended to the functional process expressed in the program would give the proprietor of the computer program copyright is a patent. There are several very serious repercussions of recognizing patentlike protection under the guise of copyright. First, it would be available without satisfaction of the patent test of inventiveness, which is a very difficult-intentionally so-test to meet. The copyright test is simply one of originality. Secondly, there would be no scrutiny by an administrative agency before the grant of protection. The Patent Office scrutinizes and evaluates every patent application, whereas the Copyright Office simply examines the technical sufficiency of the registration application without passing any qualitative judgment on it. Third, patent protection lasts 17 years. Under Senate bill 597, a copyright lasts for the life of the proprietor plus 50 years, or approximately a period of 75 years.

It is quite possible that recognizing patentlike protection for computer programs would very seriously stultify the programing art, which is, by and large, a derivative art based on fairly well established and commonly used mathematical and logical principles. It is also quite possible that the recognition of monopoly power over computer programs would create rather serious problems of vertical integration in the computer industry. Dominant hardware manufacturers might gain control over a major share of the copyrights in the software or computer program area and impair competition.

On behalf of the ad hoc committee, I would like to urge this committee to clarify the proposed copyright act so that if copyright protection is to be granted to computer programs at all, and it is far from clear that any such protection is needed, that the copyright monopoly extends solely to duplication or replication of the program, and does not incorporate the art, process or scheme that is fixed in the program. Thank you.

(The complete prepared statement of Mr. Miller, above-referred to, follows:)

STATEMENT OF ARTHUR R. MILLER, PROFESSOR OF LAW, UNIVERSITY OF MICHIGAN

The bill currently before the Senate (S. 597) to revise the copyright laws of the United States contains a number of provisions relating to computer programs and the use of copyrighted material in computer operations. The Ad Hoc Commit

tee on Copyright Revision, whose members represent a broad cross section of disciplines, academic institutions, and all levels of American education, has studied this bill intensively and has come to the conclusion that several of its provisions touching on computer use and computer programs may seriously hamper the educational and technological progress of the nation.

I. COMPUTERS AND EDUCATIONAL USES

The statement of Dean Fred S. Siebert of the College of Communication Arts of Michigan State University has shown the impact of the revision bill on the use of computers in instruction and research and suggested an amendment to the current bill on behalf of the Ad Hoc Committee. Although my statement will deal primarily with the question of copyrightability of computer programs and the need for some type of corrective device in the new copyright act that would prevent it from being rendered obsolete by the movement of computer technology, a brief recapitulation of the weaknesses of the bill's exemption provision (section 110) insofar as computer uses by education are concerned seems appropriate. The basic difficulty with section 110 of the revision bill as it now stands is that it eleminates the traditional exemption for schools and libraries and substitutes exemptions for computer operations with copyrighted works that are nominal rather than real. The "face-to-face" exemption in section 110 (1) does not seem to apply to computer-assisted classroom performances or displays at all. The "transmission" exemption in section 110 (2) permits the performance of a nondramatic literary or musical work or the display of a work by means of a transmission by an educational organization but only if (1) the performance or display is part of the school's systematic instruction, (2) the radius of the transmission is not more than 100 miles, (3) the transmission is made primarily for reception in classrooms, and (4) the time and content of the transmission are controlled by the transmitting organization and "do not depend on a choice by individual recipients in activating transmission from an information storage and retrieval system or similar device, machine, or process." The practical effect of section 110 (2) is to destroy any exemption for computer-assisted instruction. It prevents individualization—use of the computer by schoolchildren at their own speeds which is of the essence of computer instruction and may be its primary advantage. Secondly, the restriction to use in a classroom runs contrary to the healthy trend of modern education to eliminate the confining limitations imposed by classroom walls. Finally, the 100 mile radius is arbitrary and apparently intended to destroy the viability of computer "networks"-a bias that is unexplained by the House Report and difficult to justify. The amendment to section 110 proposed by the Ad Hoc Committee and the codification of the principle that an injunction should not be available against educational users should strike a more reasonable and socially desirable balance between copyright proprietors and educational users than does the revision bill.

II. COMPUTER PROGRAMS

The revision bill's expansive definition of copyrightable matter, which includes practically everything that is an original work of authorship and "fixed in any tangible medium of expression" (section 102) appears to extend copyright protection to the broad range of computer instruction and command devices that fall within the rubric "computer programs." As an original proposition, it seems highly dubious that computer programs should be accorded copyright protection. Programs are algorithmic plans for guiding the activities of machines to achieve functional results-operating printing presses, manipulating inventory, and recording payroll data-and are thus clearly distinguishable from the traditional subject matter of copyright protection-books, plays, and works of art. Moreover, there has yet to be any penetrating discussion of the implications of extending copyright protection to computer programs, as is evident upon examining the transcripts of the meetings held by the Copyright Office and the hearings held by the House subcommittee. Nor has there been any serious effort to determine whether programming economics need the stimulus of an artificial monopoly or whether the advantage secured by staying ahead of competition and keeping a program "private" is a sufficient incentive to continue the tremendous growth of the computer industry. For these reasons the Ad Hoc Committee adopted a resolution stating: "That a computer program (i.e., the instruction to the computer as distinguished from the substantive data stored in the computer) should not be subject to copyright."

If any copyright monopoly is to be extended to computer programs, the key question is whether the copyright should protect against the use of the program in conjunction with the computer to achieve the functional goal for which the program was designed. The question must be answered in the negative. Copyright should not embrace the process or scheme embodied in the program but must be limited to a prohibition against the program's duplication. A firm holding a copyright on a program ought not to have a "property" right in the mathematical, logical, or manipulative algorithms in the program that achieve a practical objective, such as operating and monitoring a printing press. Yet the broad definition of the proprietor's exclusive rights in section 106 of the bill could be read to extend protection to the process embodied in the program.

The vice of granting a copyright to computer programs that would cover the process is that its effect would be to give programs a level of protection similar to that accorded by patent, but without the safeguards and limitations that surround a patent grant. Traditionally, monopoly of systems, schemes, and processes has been granted only under patent, and only after proof to the Patent Office that there has been real "invention." Moreover, the monopoly lasts only 17 years. Copyright of a computer program would be available without regard to true inventiveness, without advance examination by a governmental agency, and, under the revision bill, the protection would run for approximately 75 years. This type of protection for computer programs might well tie up and inhibit the development of the computer program field. To some degree every program developed today is a "derivative" of the programs written for the first computers made in this country. Thus, had there been copyright protection for programs in the past, had programming been carried out under the constant threat of actions charging infringement of existing copyrighted programs, it is doubtful whether the recent growth of programs and in programming techniques would have been possible. The revision bill should be adjusted to make it clear that copyright does not extend to the process embodied in or the use of computer programs.* If any copyright on computer programs is to be recognized, it must only give the proprietors the right to duplicate the instructions in the program themselves and the right to prevent others from doing so. Even this right of reproduction must be carefully restricted, however, in order that the public's ability to employ the process contained in the program is not compromised. The power to control or bar duplication of the program cannot be allowed to prevent the use of the program in a computer.

III. THE NEED FOR A CORRECTIVE MECHANISM IN THE ACT

Probably no industry in the United States is undergoing as radical and rapid a change as is the computer industry. In the future, the computer will yield startingly unique developments in education, library science, behavioral research, the various biological and natural sciences, and numerous other fields. Almost certainly, statutory regulations adopted today that purport to strike permanent balances between creators and computer users will be rendered obsolete or function in a negative or socially undesirable manner tomorrow. Despite these risks and despite the admission in the Report accompanying the House revision bill that "the problem of computer uses of copyright material" was "touched on rather lightly at the hearings," the bill sets out comprehensive and restrictive provisions on computer uses without providing any mechanism for future revision. The infrequency of copyright revision and its consequences are well known. One merely has to examine the experience under a number of the provisions in the Copyright Act of 1909 to see how illusory the hope of revision is and how useless ancient provisions can become.

Some procedural device should be inserted in the revision bill to supply an element of self-correction; otherwise the new statute will be obsolete on the day it is enacted. There are a number of possible mechanisms: (1) the establishment of an administrative body with appropriate delegated power; (2) the creation of a body authorized to amend or add to the legislation subject to

If any statutory monopoly on a program's process is deemed necessary to encourage the programming fraternity, patent appears to be the appropriate vehicle. The question of patent protection for computer programs is being studied by the Patent Office, which recently published a set of "guidelines" for passing on applications for computer-program patents under the existing Patent Act. However, the Report of the President's Commission on the Patent System, p. 13 (November 17, 1966), recommends against patent protection.