Mr. KARP (continuing). As an example of what we are talking about; not for the record, but for the committee files. I did not mean to suggest it be printed in the record.

Obviously, the impact of machine copying on authors and publishers is vastly greater than the impact of copying by hand. It takes 32 minutes to copy a printed page. At least it took someone that long to copy page 63 of Copyright Study No. 29. We have heard many witnesses talk about how in the good old days we used to write by hand. So we forced an office boy to copy the page. It took 32 minutes. At that rate it would take him 10 hours to copy 20 pages by hand. On the Xerox machine, which we also test ran for this, it takes 4 minutes and 42 seconds to copy these 20 pages, as compared to 10 hours by hand.

At that rate, it would be ludicrous for authors and publishers to have been concerned about the impact of such hand copying on the sale of their works and absolutely inane for them to have sued anybody for copying by hand.

The rationalizations offered to condone unauthorized copying by machine are completely invalid. I have discussed them in my statement. I will not bother reading them. The point is this. These rationalizations ignore the fact that the value of a work can be eroded as easily by a multiplicity of unauthorized use as by one concerted infringement.

If an individual or an institution borrowed a copy of a book from a library and ran off 100 copies on a Xerox machine, or any other method, 100 copies at one time, and sold them or gave them to me and to 99 other people, he would be guilty of infringement. If he ran off the hundred copies one at a time, as each of us ordered them, and gave them or sold them to us, it would still be an infringement. In either event a hundred copies would have been made and placed in the hands of a hundred people without the author's consent, and without any payment to the author or publisher.

Now suppose that I and the 99 other persons each borrowed the same copy of the book from the same library and each of us made a copy for himself on a Xerox machine. The result is the same. One hundred copes have been made and placed in the hands of a hundred people without the author's consent, and without any payment to the author or publisher.

Now I had not intended to belabor the Xerox machine; we don't object to copying technology, we welcome it. It so happens that when someone makes a superior product they must suffer the penalty of being cited an example, and I am only using them as an example. There are many other copying machines or methods ranging from mimeograph and office model offset printing machines to a variety of photographic and microphotographic processes, and combinations of these systems. Any of these could be used under the proposed exceptions to the copying right that you have been asked to put in the new law.

Let me give you one example of what could be done by schools, universities, or libraries if they were permitted to make copies of a copyrighted work without the author's consent. Here is a book of poems by Randall Jarrell. It was published by Atheneum Press in 1960 and it won the National Book Award. It contains 64 pages of poetry, and it sells for retail price of $3.75. It is possible to make 50 copies of a

page from this book at a total cost for materials of approximately 43 cents-43 cents for the entire 50 copies. This can be done by putting the book on a Xerox machine (the office model), and putting a mat into the machine in the bottom where the copying paper is. These are the mats.

If I may, I would like to hand some samples to the committee. This mat is just put in the bottom of the machine, just as paper is put in. You push the button on the Xerox machine, and in less than a minute the page you want to copy is copied onto this mat. You then place the mat on a multigraph machine and they sell for as little as $1,500 and they are in many law offices, business concerns, and other institutions. They are in schools now, and, at the lower price they will come in the future, they will be as common as the mimeograph machine; and they are much better. In 2 or 3 minutes, the machines run off-the smallest machine makes 90 copies a minute-50 copies of the page.

Also for the purpose of illustration, we made a copy of page 7 of the revision bill. This is the mat bearing a copy of the page. You can tell, because it is pink on the back. This is the mat that comes out of the Xerox machine when you copy from page 7. With all due respect to the Government Printing Office, I think we have a better copy than the printed copy.

Mr. POFF. How does the cost compare?

Mr. KARP. Eight-tenths of a cent per page, in this instance. The book costs-I think it is 64 pages $3.75, several cents a page. These are the pages that come off the multigraph machine. If I am not making the process clear, you are making a master which you used to make by typing on a stencil or which a printer makes by setting type and putting it on a mat. You are making that directly by use of the Xerox machine. This is a great improvement over offset printing. You are doing what the Chinese book pirate does today on Taiwan with a camera. He copies by camera the Encyclopedia Britannica, takes the pictures and puts them on plates and runs off copies. You can do it with the Xerox machine. So far as the cost is concerned, it has to be lower. The reason is that the copyist does not have to compose the type; he does not have to do editorial work; he has no advertising, no shipping expense, no royalty. All he has to do is take the final result of the publisher's investment in preparing that book-a published copy-and go to the last step and just run off copies from it. Running off the copies is the cheapest part.

I have some additional figures which I can give you, if necessary. As I mentioned, the total cost of the 50 copies would be 43 cents, or eight-tenths of a cent a page. These costs could be reduced, if less expensive paper is used. We made our computation on the basis of fairly good paper, which costs about $2.50 a ream. The mat costs 10 cents. It costs 8 cents to run it through the machine, approximately. If you make more copies-and this is one of the most important facts, the more you make the cheaper it gets-the unit costs can be reduced even more. Fifty copies of this entire book could be produced for a cost of about 55 cents per copy, compared to the retail price of $3.75. That is direct cost, not counting labor, because it depends on who is running the machine. If it is a student working for a dollar an hour, you have to figure on that basis. If it is someone standing behind a counter not doing other work at the time, he could do it.

Mr. KASTENMEIER. That does not compare binding.

Mr. KARP. That is true, but the question is how often can you serve the same purpose without binding.

Mr. KASTEN MEIER. You would be better to compare it with a paperback, would you not, in terms of cost?

Mr. KARP. Yes.

Mr. POFF. May I interrupt?

Mr. KASTEN MEIER. Yes.

Mr. POFF. Would the practice that you have just described be an infringement under the present law?

Mr. KARP. Yes, it would, sir.

Mr. POFF. It would be an infringement under the bill as it lies before us today?

Mr. KARP. Yes.

Mr. POFF. The point you are making is that it would not be an infringement if the amendments which have been recommended in previous testimony were adopted.

Mr. KARP. That is correct, sir. For certain types of copying. In justice to the proponents of these proposals, I don't suggest that they claim that any and everything can be copied this way, but certain things can be.

I come to that in a moment, to illustrate how it could be used under the proposals. I am not suggesting, for example, that Mr. Rosenfield has come before you to ask for the right to copy everything. He has not. But I will address myself precisely to the type of copying that I think could be done under those proposed exemptions.

Now even more important than the cost per copy is the cost it would require to make a portion of the copy of this book, because a portion of a copy can often serve the demand for the whole copy. If you can get half of a book of poetry for 30 cents or 25 cents, or a third of it, you may never buy the whole book, especially if you are using it for studying and teaching. Obviously, the making of unauthorized copies of an entire work deprives the author of royalties and the publisher of sales, but the making of unauthorized copies of a portion of a work can be equally, and sometimes even more, harmful. In any event, it also unjustly deprives the author and publisher of fair compensation for the use of something of value that they have created.

I have given a few examples in my prepared statement. I will read one of them now. One of the greatest potential dangers of permitting copying beyond fair use is the impact it would have on poets, novelists, short story writers, essayists, and others who derive a substantial portion of their income from granting permissions to others to use a poem or a story or excerpts from it in anthologies or textbooks. Licenses for such uses are given on a nonexclusive basis, and the same poem or excerpts from it may be used in several different anthologies or textbooks. Thus, many people may be reproducing, under license, copies of a single poem or a portion of it, or a story. Short excerpts from novels and other books are also used in anthologies and in textboks for teaching purposes, and the right to make such uses bring substantial income to the author and the publisher.

Now aside from anthologies and textbooks, I should point out that it is not uncommon for authors to earn income from uses of portions of their work. When a man writes a novel, he can be paid for the

right to print an excerpt from a chapter, in a magazine like the New Yorker or Harpers. He does not have to sell the whole book to make money, only a part of it. Sometimes the money from the publication of a single excerpt in a magazine may bring more than from the whole book, which may be a total failure. Excerpts are published in Reader's Digest and other digests, and serialized in newspapers.

The idea that you only make your money from selling an entire work does not gibe with the economy of professional writing and publishing in this country. Under the cover of an exception that transcends fair use, and using either the process I have just described or typewriter-stencil-mimeograph-and, by the way, there are also processes for transferring a direct image onto a mimeograph stencil-the Xerox machine, or any other device, a school or a teacher or a school system could assemble its own anthologies. Remember, that an anthology does not have to be bound or even produced at one time. If it is created in installments, a poem at a time, a short story at a time, it can serve the same purpose in the classroom as a bound anthology-the purpose of communicating or using a poem or short story or excerpt for teaching purposes.

Copying from an author's work in this manner would diminish the demand for anthologies, stifle the publication of new anthologies, and thus drastically reduce his income. It would also mean that the author's work was being used in precisely the same manner that other people were paying him for the right to use them in textbooks and anthologies. If we pay for the other tools used in teaching, we certainly should pay for this primary and most important one. The injury to authors would be severe. A large portion of all anthology sales are made to schools, and a substantial part of a poet's income is derived from the fees for such uses.

Here is an anthology called "Adventures in American Literature," published by Harcourt-Brace. I would like to offer this to the committee. Here, for example, are several pages of poems by modern American poets, including Richard Wilbur, who also had books of verse published. Here is one entitled: "Things of This World." One of his poems from a book like this is in that anthology, and he has been paid a fee for it. It is a short poem. As you look through that section, you will find that all these poems are a page or less, and their authors are paid for these uses. If you look through the anthology, you also find that they form part of a teaching lesson. This is a textbook anthology. It says, "After you read the poem, analyze it. What do you find?" I ask you, what is more important as a contribution to the education of the student who is reading this-the poem itself, or the little paragraphs of comments by the author of the text or the teacher in the classroom, at the end of the poem?

A substantial part of the poet's income is derived from the fees for such uses. Indeed, the greater part of his income may be derived from permissions to use his work in these anthologies and textbooks rather than from sales in the original trade edition. An attorney who specializes in the publishing field recently told me of a poet, a close friend of his, who derives 75 percent of his writing income from the fees paid for permissions to use single poems or excerpts in anthologies such as that, sold primarily to schools. He also informs me that it is likely that if his publisher clients-the publisher shares

52-380-66 pt. 3 25

in this sort of subsidiary income-were deprived of income from permissions such as this they probably could not afford to publish the original trade edition of poetry, because it is so expensive and the sale is limited. They can bring the public the trade edition because other people pay them for anthology rights. The author can earn some income for writing these poems, because people pay him money for the right to publish parts or excerpts in other areas, and primarily for education.

I would like to add one other comment. Some educational witnesses have told you that free use of literary works in schools would stimulate their sales elsewhere. But the destruction, in this instance, of the most substantial market for some writers, the educational anthology and textbook, would deprive authors of a greater portion of their income and possibly destroy all sales in the trade market as well.

Moreover, it is commonplace in all fields of copyright for an industry to tell the same story to authors. "We help publicize your work in other media." "We play your records," say some people, "therefore we should not pay you. Let the record manufacturer pay you."

The record manufacturers say, "If we did not make your records, they would not be performed on radio and television. And look at all the money you make on radio and television."

The radio and television broadcasters say, "If we did not play your records, who would buy them?" If the author-and this is true all down the line-if the author accepted this principle, "We help you, so we should not pay you," nobody would pay him.

We are not proposing that the new copying techniques and devices be prohibited from disseminating the works of American authors and publishers, but we do believe that authors and publishers have the right to be paid when their works are communicated by these devices. Money appears to be the principal stumbling block. Methods for granting permission to make copies can be established. But those who advocate the various exceptions to the copying right simply object to paying authors and publishers. Inconvenience may be the rationalization, but dollars are the reason. They are willing to buy or rent copying machines. They are willing to pay for copying paper and the other copying tools. They are only unwilling to pay for the one essential ingredient which all these other materials serve to communicate, and only exist to serve that purpose-the author's work. This is a paradox.

The whole purpose of spending money to make copies of copyrighted work is to obtain what the author has created. This is the thing of value, not the paper, not the whirring machinery. And yet this is the one thing for which they object to paying. We do not contend that all copying by machine or otherwise is prohibited. We do not suggest that every time a teacher copies lines from a poem or paragraphs from a book for her class she is necessarily guilty of infringement. On the other hand, we do not admit, concede, or believe that all unauthorized copying is permissible. The doctrine of fair use applies here as elsewhere. You have been told that fair use is too difficult for teachers to understand and comply with, and that a more definitive guideline is needed. I would like to make these final

comments.

1. No one has come forward with a detailed and definitive guideline. No one has.