identity of the registrant, the copyright expiration date, the serial number, the identity of the maker, the copy number and the day of copying are incorporated. Policing the system for payment

Section 117 provides for royalty payments for each and every authorized reproduction. Payments are made directly to the registrant, along with a listing of the authorization records under which the copies were made. Under section 117 (b) (5), the same authorization records which are required to police the system against propagation of errors in text are also used to prepare the statements of accounting to registrants.

Policing the system against infringements

The authorization records required for strict accounting also permit unique identification of every authorized copy. The absence of a complete authorization record on a made-to-order copy of a work registered under this section is prima facie evidence of infringement. Since no two authorization numbers can be identical, the appearance of copies with the same authorization record is also prima facie evidence of infringement.

Provision for blow-back copies

It is recognized that some users may wish to convert authorized copies from forms requiring the aid of a machine or device to legible form. Section 117 (k) (3) provides for such a situation. Under it, no fee is required for preparing one, and only one, legible version of the text in a non-legible authorized copy. Further reproduction from the authorized copy or from the legible version is not authorized.

Infringement

Section 117 (f) (1) brings copyrights created by this section under the provision of chapter 5 for the purpose of policing exclusive rights.

In a system of permissions and payments involving minimal fees, actual damages for individual infringements are inconsequential. The primary value in infringement provisions is their deterrent power. Granting of an exclusive right for the registrant to sue for an infringement under this section would not provide an effective deterrent. Clauses (2) and (3) of section 117 (f) authorize the Register of Copyrights to police infringements as the beneficial owner, to sue for statutory damages; and, to pay damage awards to the general funds of the Treasury.

The minimum statutory damages for infringement under this section are reduced to $25 in the case of each wilful infringement and $10 in the case of each unwitting infringement. It will be noted that the minimum penalty for infringement is 100 times as great as the license fee provided in section 117(b) (3). Effect of failure to register under section 117

Copyright owners are not required to register their works under the provisions of this section. If copyright owners elect to distribute their literary works to the public in forms that are illegible without the aid of a machine or device, they must, under the provision of section 117(f) (4) enroll their works strictly in accordance with the provisions of the system or lose the right to sue for infringement of the exclusive right to reproduce and disseminate said works in copies. Effect on negotiated licenses

Registration of a work under section 117 limits the duration of the exclusive right to make and distribute the work in copies to 20,000 days (54.757 years). Otherwise, it has no effect on the rights of the copyright owner to license the making and distribution of copies in legible form,

Registrants are free to negotiate the fee payable for each reproduction in forms that are not legible, or to negotiate blanket licenses therefor. An agreement to waive the requirements for keeping authorization records on non-legible copies, however, would be the equivalent of abandoning the policing system provided in this section. It would impair and disrupt the effectiveness of the policing system, making it impossible to distinguish between unauthorized copies and copies authorized under negotiated license.

[Exhibit 1 to Statement of N. Goodwin-Apr. 6, 1967]

Title: Automated Information Storage and Retrieval: Permissions and Payments.'

Author: Norton Goodwin."

ABSTRACT

In the absence of an effective deterrent, the publishers' competitive advantage in applying automation to printing may be more than offset by the unauthorized application of the same sort of technology to automated storage and selective retrieval of his publications. The public interest in securing information copies of enrolled publications upon the payment of a license fee is considered, along with the public interest in keeping proprietary rights in literary works separate and distinct from proprietary rights in automation techniques. In a system of permissions and payments, the most effective deterrent to unauthorized copymaking is shown to be a requirement for one-for-one correspondence between each identified copy and the authorization record on the basis of which a proper fee was paid, with the burden of establishing proper identity and proper correspondence upon the maker and/or holder of such copy. It is shown that automation technology can provide documentary proof of copymaking authority as a byproduct of systematic accounting for fee payments. The requirement for legislative implementation is stressed. A statutory system for compulsory licensing of enrolled literary works is compared with existing law governing musical recordings.

TEXT

I am delighted with this opportunity to address members of the printing and publishing industries at an Institute on Management of Automation, and to talk about systems of permissions and payments for the use of intellectual property in the new information storage and retrieval technologies. Since it is idle to discuss payments for property which, under our present copyright laws, can be "lifted" with impunity, the real subject of my talk is statutory systems of deterrents to unauthorized copymaking. It is a subject of major significance if the public interest in sustaining commercial publishing activity on the one hand is to be reconciled with the public interest in getting automated access to published information on the other.

First of all, let me say a few words about why I have been concerned about preserving the kind of information services, in content, if not in form, to which I now have access. In our firm, we subscribe to no less than nine independent publishing firms for current federal tax information services. Of these, six are looseleaf services which are updated weekly. As a publisher associated with the Society of Photographic Scientists and Engineers, I have had considerable experience with computer storage and retrieval of information, as well as with computer utilization in composition. I have been aware for some time of advanced informtaion-transfer technologies which, in my opinion put many kinds of publications in financial jeopardy.

Again, as a lawyer, I am aware that opposing Internal Revenue Service counsel has computer access to information to which I do not have access. I am not at all concerned by the fact that memoranda from the Chief Counsel's Office, unpublished rulings, et cetera, are inaccessible to me, but I am deeply troubled by the advanced technology now being used to put this information at the disposal of opposing counsel. How long, I wonder, before government agencies begin automating their entire law libraries in the same way that they are now automating their inhouse reports and memoranda? As a publisher of scientific and technical literature, I wonder how long before these new storage and retrieval technologies in the hands of government and of industry subscribers will destroy the revenue basis for the scientific publications for which I am responsible? Technological advances due to automation are affecting publishers in two different ways. On the one hand, the use of computers for line-justification, hyphena

1 Text of talk at Institute on Printing and Publishing: Management of Automation, January 16, 1967. 282 Connecticut Avenue, Washington, D.C. 20006, Tele: 202+783-4100.

tion, and even for page make-up, is beginning to produce better composition, faster and more economically than ever before. This kind of automation is entirely under the control of management.

Automation can also be used by third parties beyond the control of management to provide faster and, in many cases, better access to the intellectual property in which publishers seek to retain an interest through copyright. It is characteristic of this other kind of automation that the particular copy that reaches the consumer is made by persons and processes beyond the control of the publisher. and is made only if, when, and as needed. Unless the Congress is able to extend to copyright owners some effective measure of control over this secondary copymaking activity, essential information services now provided by commercial publishers will no longer be available.

It is encouraging to learn that the publishing industry is beginning to recognize the nature and extent of the threat represented by automation. It is encouraging to find that serious studies on the economic impact of the new information technologies are beginning to point to copyright as the most significant factor. In the interest of time, I am going to assume that as publishers, you have read, or shortly will read An Economic Media Study of Book Publishing copyrighted in 1966 by the American Textbook Publishers Institute and the American Book Publishers Council, New York. The study documents the impact of automation on publishing with a wealth of factual information about what is now going on in the way of replication practices, particularly in the federal government and in schools, and about the new information technologies. The importance of copyright revision is emphasized in each of the first three papers. It is inferred in the fourth.

In his paper on An Economic Analysis of the New Information Technologies, Professor Jesse Markham of Princeton University observes in the matter of copying part of a book . . . ". . . the copyright to a book should not obstruct extraction of the maximum use of a book, but neither should the increased incentives to use books in ways different from past patterns lead to emasculation of the property right. The conflict of seller and buyer can be resolved here, as can most other buyer-seller conflicts, through the pricing system." As regards the new information-transfer technology, he concludes: ". . . while it is apparent that the copyright should not obstruct the economical use of the new technology, it is equally apparent that the new technology should not be made the means of riding roughshod over the rights of authors and publishers in the intellectual property that they create and prepare for us by the public."

Since it is obviously impractical to price the right to copy a literary work except on the basis of use, Dr. Markham's reference to the pricing system assumes the existence of effective statutory deterrents to unlicensed copymaking. There are none at present.

In his paper on Federal Programs and Commercial Book Publishing, Professor Lowell Hattery of American University concludes re copying: "It seems, therefore, that the interests of publishers will be best served by participation in the new media forms and their reproduction and by developing systems of licensing to copy with appropriate charges."

Dr. Hattery is fully aware of the impracticality of licensing literary works without adequate copyright protection. It is fair to conclude that Dr. Hattery is suggesting amendments to the copyright laws that would make "systems of licensing to copy with appropriate charges" operational.

In discussing Information Technologies with Broad Implications for Book Publishing, the Arthur D. Little Inc. study on the Impact of Technology on Publishing observes: " the capability for copying and duplicating copyrighted books presents perhaps the most immediate and readily definable technological threat warranting consideration in the revision of copyright law. However, there is a broad area of technological development outside the copying/ duplicating field that can be expected to change significantly the pattern of information dissemination and use in the future, and consequently, this area might appropriately be considered if copyright law revisions are to achieve long-standing validity."

There is no question but what the Congress must supply specific means whereby the public interest in automated access to information can be reconciled with the public interest in securing proprietary rights in literary works if the Congress expects to come up with "copyright law revisions that are to achieve longstanding validity." On the other hand, the feasibility of dealing in advance with the new capabilities for copying and duplicating hard copy books as a separate problem is open to substantial doubt.

The particular bone of contention in copying all or parts of published books is the question of "fair use." The right to free access to copyrighted books in a free library is beyond challenge. Copying of library materials under certain conditions has long been recognized by the courts as a valid and reasonable exception to the exclusive rights granted under copyright. This kind of copying has been held to be a "fair use" in a long line of judicial decisions.

Congress has already considered the technology threat posed by the capability for copying and duplicating copyrighted books. Dr. Hattery has played a useful and important role in documenting the "fair use" problem through an Institute on Copyrights and Reprography. Voluminous testimony on the subject, in addition to commissioned studies, were presented at hearings before the House Committee on Copyrights over a ten-year period. In the bill finally reported out by the Committee in the last Congress, the "fair use" exception is spelled out in the copyright statute for the first time. The provision follows:

“107. Limitations on exclusive rights: Fair use

Notwithstanding the provisions of section 106, the fair use of a copyrighted work, including such use by reproduction in copies or phonorecords or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching, scholarship, or research, is not an infringement of copyright. In determing whether the use made of a work in any particular case is a fair use, the factors to be considered shall include:

(1) the purpose and character of the use;

(2) the nature of the copyrighted work;

(3) the amount and substantiality of the portion used in relating to the copyrighted work as a whole; and

(4) the effect of the use upon the potential market for or value of the copyrighted work.

It is manifest that by incorporating what was formerly a judicial doctrine into a specific statutory limitation on exclusive rights, the Congress would be contributing nothing in the way of a deterrent to unauthorized copymaking.

In my opinion, there is no essential difference between storing literary works in machine encodements or graphic microform and making copies on request and the activities of a librarian who stores published works in the form of hard copy, and supplies photoduplicates on request rather than let the original copy leave his shelves. If a copyright law can be devised which reconciles the public interest in automated access to literary works stored in technical forms with the public interest in protecting proprietary rights, it is quite clear that we will have at hand means for preventing unauthorized copying from hard copy, as well. Publishers will be in a position to close the "fair use" loophole by providing the right to make licensed copies of the same work via automation. We would do well, it seems to me, to investigate the feasibility of a statutory licensing system for "recorded" literary works, even if we were primarily concerned with protecting Lard copy.

In a system of permissions and payments involving the collection of a small fee for each authorized reproduction, the execution of a deterrent penalty must be as certain and as automatic as practical, but the penalty should not be so great as to invite judicial intervention. I suggest that the penalty for infringement apply atuomatically wherever a person is found to possess or to have made a copy of an enrolled work that does not show on its face a record of the transaction under which it was authorized. This would place the burden of showing compliance with the statutory licensing provisions directly on the user, and this is the exact inverse of the way copyright law now operates with respect to hard copy.

In my opinion, the best evidence that a particular copy has been authorized is a printout on its face of its own unique authorization record. An operational system providing such evidence would require a one-for-one correspondence between each authorized copy and a record of each transaction in which an authorized copy was made. Records of copy-making are going to have to be kept, in any event, for the purpose of making payments and for identifying active, as opposed to inactive, files.

In my opinion, an optimized system for licensing the making of a copy of all or part of a literary work upon the payment of a small fee fixed by statute must be designed to afford maximum deterrents with a minimum of record-keeping osts. For this reason, I feel the statute should be drafted with automated recordkeeping specifically in mind.

In drafting an automation-oriented statute designed to deter unauthorized copymaking, care must be taken that all operational instructions can be executed on the basis of data in hand. This calls for specification of an efficient format for the notice of copyright in which the unique identity of each work, the identity of the payee, and the expiration date is to be given. To be efficient, the format must recognize that payee and expiration information are part of, and not additions to, the identification of the work.

Similarly, the format for the authorization entry must be specified with a view to storage economy. In addition to the identity of the work, the authorization entry must include fields for the identity of the authorized copy-maker, the execution date, and for the serial number, should more than one copy of the work be executed on that day.

Record-keeping cost considerations require that the requisite fields be kept as few and as small as practical. For this reason, the chronological data records that must be made if the statutory deterrents are to be upheld as constitutional should be kept in terms of day numbers. Article 1, Section 8 of the United States Constitution grants Congressional power to create exclusive rights for limited times only. Use of day numbers permits reducing storage requirements for chronological records by a factor of three. In a system requiring billions of chronological records, such an opportunity cannot be disregarded. I have included in a handout, a discussion on Day Numbers for Recording Chronological Data, which I wrote for the Independent Tracking Coordination Program Bulletin for December 30, 1966. It includes computer subroutines for conversion from day numbers to Gregorian calendar date and back.

There is good precedent in United States Copyright Law for special licensing systems to accommodate technological advances. The factors that led to the enactment of a compulsory licensing system for musical recordings are documented in the Statement of the Record Industry Association of America, Inc. on pp. 667 through 704 of the Copyright Law Revision Hearings, 89th Congress, First Session. It will be noted that the particular technological advance that had to be dealt with in 1909 was the automation of music by player-piano rolls, but that the system then created has since accommodated all media for recording music subsequently developed. There is no reason to believe that a compulsory licensing system for literary works recorded in microforms and machine-readable encodements now known should not be as serviceable in the case of forms later to be developed.

In the context of the present statutory licensing system for musical recordings, the word, compulsory, in effect. means nondiscriminatory. The compulsory system does not exclude negotiated licensing of the right to make copies. In fact, most record-making is done under negotiated licenses. In the case of musical recordings, the compulsory licensing provision prevents exclusive licensing of the right to make recordings from musical compositions. It is designed to prevent undesirable monopolies. It assures C the right to make musical recordings of A's work upon payment of the fee fixed by statute in any case where A gives reproduction rights to B, regardless of how little B must pay.

In the same way, a compulsory licensing system for "recordings" of literary works would protect small publishers and small information-retrieval activities from monopolistic practices. It would prevent tying published information to proprietary technology. There is a public interest in being able to enjoy advances in information storage and retrieval technology as they are developed, that is directly comparable to the public interest in being able to enjoy new techniques in recording music, without having to wait for copyrights to expire. In this sense. then, compulsory licensing does not imply involuntary expropriation of preexisting rights. In this sense, a compulsory licensing system for literary works would be restricted to the works that are voluntarily enrolled. Application for copyright on a book or periodical would not constitute enrollment in the licensing system any more than application for copyright on a sheet of music subjects the composition to compulsory licensing under present law. Hard copy of works that had been enrolled could be distinguished from copy subject to "fair use" copymaking by the appearance in a conspicuous place of the statutory notice in which is found the expiration day number, N. If this number, N. is found to be greater than the current day number, M. then making a copy without a license should automatically constitute an infringement. In terms of automation, the statutory license fee requirement for making a copy would be invoked by the machine instruction: