Before, and to some extent after, the invention of the typewriter, all manuscripts were just that — handwritten documents. Along with the author's notes, this allowed us to see how the author actually created the work. One could actually see the work evolve, be amended, changed, etc. — for that is really the true value of a manuscript. It allows you insight into the wrok not otherwise attainable (and, by its very nature, this type of manuscript is unique).

Then, with the invention of the typewriter, we had the "typescript," often amended in the author's handwriting and usually done in several drafts. This still allows us to see the evolution of the work as it progresses (though not as clearly). The typewriter is also able to create a second sheet, or carbon, so now we have a "ribbon-copy" typescript and the potential for a "carbon-copy" typescript.

Now, with the invention of the computer and word processing, we have the dreaded printout. So we have progressed, if that's the term, from a unique handwritten document, which gives us a clear insight into the author's creative process, to a sterile, clean printout that can be produced in potentially limitless numbers and that tells us nothing about the author's creative process. The only thing that makes the printout potentially of value is provenance. For example, the "setting-copy," with a signed letter of provenance from the author stating that this manuscript was produced by him or her for submission to the editors at the press, etc., along with the editorial corrections that may be present, would constitute impeccable provenance, showing that this was a contemporary working copy and not run off ex post facto.

Now, with the elimination of even the "setting-copy," the future of manuscript collecting and literary scholarship is truly in danger. If authors must use word processors (and I can see that they will) then for the sake of posterity they should be sure to make printouts at all stages of the work-in-progress. Before changes are made on the screen, a hard copy of the original version should be made. All drafts should be labeled so that the order can be determined. Keep all notes. Sign and date all drafts and notes. Keep all notes and drafts in order and together. Last but not least, time should be taken to write a note or letter of authentication or provenance, and it should be as detailed as possible. If the author has anything he would like future scholars to know about the work, it should be included in the letter. (Why let someone a hundred years from now try to tell people what the author "really" meant when the author can do it himself?) The items should be stored in a safe place, one that's not too hot or too damp. Someday the author (or the author's relatives) may be very happy these things were done.

The value of the manuscript depends on many factors: the importance of the author and the work, the form and condition the manuscript is in, and the scarcity of manuscript items by that author on the market. But unless the author makes an effort to preserve manuscripts in a thoughtful and useful manner, he or she cannot profit from them — and neither can our culture.

They are among the first types of book collectors to stop "following the flag" (i.e. buying the first edition of a book printed in the author's country of origin regardless of where the "true world first edition" was published). For a modern example, a number of Stephen King's books have been published in England, one in France, and another in Germany, all of which are eagerly collected today.

It is also true that such collectors will no longer refuse to buy a book club edition when that edition is the true first edition, such as Isaac Asimov's Forward the Foundation for which the "true world first edition" was the BCA (a British book club) edition.

No wonder then, that science fiction, fantasy, and horror collectors are among the first collectors to recognize and adopt the somewhat controversial view that Advance Readers Copies can and often do constitute the true first edition of many modern books.

Now let us clearly define what we are discussing. We are referring to the so-called "Advance Editions," that is, complete books produced and distributed (before the trade or limited editions) by their publishers, usually to promote the sale of the trade editions either by sending them to bookstores, or by giving them away at conventions of the American Booksellers Association, or by putting them in the hands of reviewers (hoping to get the trade edition favorable publicity).

Often, but by no means always, advance editions are bound in soft covers (wrappers). They are variously designated by such terms as: "Advance Readers Copy," "Special ABA Edition," "Special Advance Copy," "Special Advance Readers Copy," "Special Edition Not for Sale," "Special Preview Edition," "Special Reading Copy Not for Sale," etc.

Under this definition, we are not discussing bound or unbound offset or photocopied copies of the authors' typescripts, or bound or unbound galleys, proofs, advance proofs, uncorrected proofs, etc. These are all parts of the pre-publication process and thus are not considered "editions" (not to be confused with "Advance Reading Edition from Proof Pages" or "Advance Readers Copy from Proof Pages," which are forms of "Advance Edition").

It is true that John Carter, in his famous book first published in 1952, ABC for Book Collectors, wrote, "Such advance copies...are naturally of interest to the keen collector...but they do not (as is sometimes suggested) represent a first or early issue in the proper sense of the word nor can the existence of 50 advance copies of a book prejudice in any way the firstness of the first edition as issued on the day of publication."

John Carter, who also wrote Taste and Technique in Book Collecting,where he accurately predicted that science fiction and mystery fiction would become highly collectible, also predicted that tastes and techniques in book collecting change and modern collectors now find it more and more difficult to ignore that fact of the existence of these advance editions and their bibliographical importance or "firstness."

I am also sure that John Carter, back in 1952, did not envision advance editions as large as 10,000 copies. I am also equally sure he never thought that the publishers would label these advance copies with such terms as "Special Edition," "Special ABA Edition," "Special Preview Edition," etc. Are collectors not to take publishers at their word and consider such advance copies "editions" — and, since they precede all other editions, in fact, "true first editions"?  The modern science-fiction collector is interested in acquiring the "true first edition" no matter what form the book may take.

I think that the main attraction of these advance editions to collectors is the fact that they are always a fraction of the size of the press run of the publishers' trade "first editions." There are bestselling authors whose books can have first printings of 250,000 to 500,000 copies. Owning the advance editions can mean the difference between having first editions that are collectible (i.e. first editions with small press runs the value of which will increase over time ), and having books whose press runs are so large that they will never have much value on the out-of-print or collectors' market.

The irony of the modern advance editions is the fact that most of their publishers have no idea that, in the most modern bibliographic sense of the term, they have inadvertently published the "true" first editions of their books (with all that implies) in a form not intended. They have in effect preempted their own first editions.

As a rare-book dealer, I am delighted when an important book, the first edition of which would have been printed in a press run of 500,000 copies, has been published in an advance edition of only 10,000 copies, making it more valuable to collectors in the long run. However, as a lover of books, a bibliophile, I think publishers should have real control over the publication of their first editions and should not produce them inadvertently.

With more and more people collecting modern first editions, far more than ever before, it behooves publishers to at least make informed decisions about how they release their first editions. With all the time, effort, and money some publishers put into the release of new titles, especially a major new release by a top author, I am sure they do not wish to unintentionally issue the true first edition in a form that does not do them justice.

Some authors really care about the way and form in which their first editions appear. I have heard more than one prominent author complain that the true first edition of one or more of their books was issued in a paperback advanced edition with promotional copy all over it.

If publishers do not wish to preempt their own first editions but still wish to produce some kind of advance issue, all they need do is print a large number of "proof" copies. "Proofs," "Advanced Proofs," "Uncorrected Proofs," etc. do not, in a bibliographical sense, constitute an "edition." They are part of the pre-publication printing process. They can be as plain or fancy as publishers wish as long as they are designated as some kind of "proofs," but not "Advance Reading Edition from Proof Pages" or the like.

The following list of books is a sample of first editions, the format of which was probably not what their publishers intended. Estimates of their current values are included. The different publishers' designations appear within quotes:

Clive Barker. The Thief of Always. Harper Collins Publishers, [London, 1992]. "Advance Copy." Limited, signed and numbered, first British and first limited edition and possibly true first edition. There is also an American "Advance Readers Copy" (which was not signed and numbered) and it is not yet known which of the advance editions was issued first. The British edition was limited to only 250 copies. Wrappers. $300.

Greg Bear. Moving Mars. Tor: A Tom Doherty Associates Book, [New York, 1993]. True first edition. "Advance Reading Copy." Wrappers. $95.

David Brin. The Postman Bantam Books, [New York, 1985]. True first edition. "Advance Reading Copy." Wrappers. Soon to be a major motion picture starring Kevin Costner. $150 & up.

Terry Brooks. The Sword of Shannara. Random House, [New York, 1997]. True first edition. "Special Preview Edition." Illustrations by the Brothers Hildebrandt. Wrappers. $255 & up.

Thomas Harris. The Silence of the Lambs. St. Martin's Press, [New York, 1988]. True first edition. "Advance Reading Copy." Wrappers. Although the book has no fantastic or supernatural elements, this book won the Bram Stoker Award for Best Horror Novel of the Year and rightly so. This book was made into a major motion picture. Wrappers. $85.

Robert Jordan. Lord of Chaos. Tor Fantasy, [New York, 1994]. True first edition. "Advance Bound Reading Copy." Two volumes. Self-wrappers. The two volumes are laid into a copy of the dust jacket for the hardcover trade edition. The deluxe version of this, the true first edition, was issued in a wooden box whose siding lid bears two engraved silver-tone plaques, one of which reads, "Robert Jordan the Wheel of Time" with the "Wheel of Time" logo, and the other which reads, "Compliments of Tor" with the "Tor" logo. Also contained in the box is a grail engraved "Garden Pewter 532 Jefferson Cup." It has been reported that as few as 10 copies of this deluxe state were issued. The number has not as yet been verified. A promotional folder is also enclosed in the box. Jordan's most elaborate edition to date and sure to be a most sought after collector's item. $600 & up.

Stephen King. Carrie. Doubleday & Company, Inc., [Garden City, 1974]. True first edition. "Special Edition — Not For Sale." The first of two issues must have the code "050" in the gutter margins of page 199. Wrappers. The first true edition of the author's first book. $950 to $1,500.

Stephen King. The Dark Tower III: The Waste Lands. Donald M. Grant Publisher, Inc., [Hampton Falls, 1991]. True first edition, first issue. "Advance Review Copy." Wrappers. Issued in shrink-warp. With an envelope containing 13 color plates for the book and with a form letter from the sales manager of the press, as issued (copies without the envelope, plates, and letter are not considered to be bibliographically complete and as issued). Must have the original plates issued with the book and not one of the plates offered later for sale by the press. $500 and up.

Stephen King. Gerald's Game. Viking. [New York, 1992]. True first American edition. "Special Limited A.B.A. Edition." This edition was printed for distribution to booksellers attending the 1992 American Booksellers Association Convention and was intended as a thank you from the author and publisher to the booksellers for helping to make King a bestselling author. A letter to the booksellers from King is reproduced on the front free endpaper and states in part, "I and the people at Viking wanted you to have this special limited ABA edition of Gerald's Game to say thanks for all the good years and good times we've enjoyed together." Hardcover. Issued in a cardboard slipcase (not issued in dust jacket). $150 and up.

Stephen King. Pet Sematary. Doubleday & Company, [Garden City, 1983]. True first edition. "Promotional Copy." The first of two "Promotional Copy" editions produced by photo-offset from King's original typescript. This first edition was supressed by the publisher and only 50 to 100 copies survived the destruction of the edition. Wrappers. This book was given away as part of a booksellers' contest held in Publishers Weekly. It is a rare case in which the author's typescript was actually turned into an advance edition by the publisher. $500 and up.

Richard Bachman (pseudonym of Stephen King). Thinner. New American Library, [New York, 1984]. True first edition. "Special ABA Edition." Wrappers. $150.

Fletcher Knebel and Charles W. Bailey II. Seven Days in May. Harper & Row, Publishers, [New York, 1962]. True first edition. "Advance Copy." Wrappers. This book was made into an important film. $185.

Dean Koontz. Dark Rivers of the Heart. Alfred A. Knopf, [New York, 1994]. True first edition. "Advance Readers Edition." Wrappers. $65.

Dean Koontz. Intensity. Alfred A. Knopf, [New York, 1996]. True first American edition. "Advance Reader's Edition." Wrappers. Small press run. $150.

Joe R. Lansdale. Mucho Mojo. The Mysterious Press. [New York, 1992]. True first edition. "Advance Reading Copy." Wrappers. $45.

Ursula K. Le Guin. A Fisherman of the Inland Sea: Science Fiction Stories. Harper Prism. [New York, 1994.] True first edition. "Special pre-publication limited edition." Limited to 1,500 copies. Distributed as gifts of the author and publishers. Hardcover (not issued in dust jacket). Note the way this book has been designated by its publisher. $125 and up.

Robert R. McCammon. Mine. Pocket Books, [New York, 1990]. True first edition. "Advance Reading Copy." Wrappers. Winner of the Bram Stoker Award. $75.

Anne Rice. Interview with the Vampire. Alfred A. Knopf, [New York, 1976]. True first edition. "Special Preview Edition." Pictorial front wrapper (with a photograph of the author on the front cover). So scarce that the author does not even have a copy. $850 to $1,200.

Anne Rice. The Tale of the Body Thief. Alfred A. Knopf, [New York, 1992]. True first edition. "Advance Reader's Edition." Wrappers. (Not to be confused with the "Proof" for the first trade edition.) $125.

One last interesting sidelight to the issue of the advanced editions: A friend of mine is a small-press publisher who specializes in horror fiction and who bought the rights to do a "Limited Fine Edition" of a major book that was to be issued by a large well-known publishing house. An employee of mine, returning from an A.B.A. Convention, brought back an "Advance Readers Copy" of this same book to which my friend had bought the "First Edition" rights. The large publishing house had had this advance edition printed from proof sheets as a sales tool to promote the sale of the trade edition. I do not think my friend had ever been able to make the publisher understand why he is so unhappy.

"Suppose the first US-advertised offering by the print-on-demand publisher on its webpage was early in 2002, but there was no order for a paper copy until twelve months later, at which time the book was printed in paper form for the first time in America. What is the year of publication? Can the book be said to exist at all in 2002 since no paper copies were printed? Suppose the book was ordered over the web, in e-form (instead of paper) and was read in 2002 by a buyer on his computer screen and wasn't printed out. Is this a publication? A first? How does this relate to the subsequent print-on-demand first paper copy?"

My writer had referenced a number of books on bibliography before writing me but said none of them addressed these questions. I suppose, because I had written other articles dealing with the effect of new technologies on bibliographical questions, he thought I could help. I gave some long thought to answering his queries. The answers would require melding our past western bibliographical traditions, which go back to Gutenberg, with the here-to-fore unthought of modern computer technologies. Unless we wish to completely do away with our previous bibliographical concepts of what constitutes a first edition, and how bibliophiles go about collecting them, we must reach some compromise with our past.

So I propose the following as possible answers to his new bibliographical questions:

The first batch of copies printed in America, bound and sold here, constitute the first American edition. (This would not be the first book to use facsimile sheets from a British edition to produce a first American edition of an out-of-copyright book. Several publishers geared to the scholarly library market have been doing this for years.) So, yes, this print-on-demand American publication can be called "the first American edition." In answer to his question whether the book could be considered to have been published in 2002, because it had been initially offered by the print-on-demand publisher in that year, but with no copies ordered or printed until 2003. I answered: The year of publication is the year in which the first printed and bound copies are first offered for sale. If the book was first printed, bound, and sold in 2003, then 2003 would be the year of publication. (Many books are advertised for sale a year(s) before they are actually published. Also, perhaps, more to the point, books are sometimes sold by subscription with orders being taken far in advance of the actual publication date, which seems to be the best parallel for the print-on-demand publications.)

As for a book viewed on screen: This would not meet the traditional definition of a published "book." The first paper copy would still be the first "book" publication. (Just as an audio-book released before the publication of the printed book does not take away from the "firstness" of the paper-and-ink copy.)

The question which wasn't asked but which does really need to be addressed for all future collectors, librarians, and bibliographers is, "What constitutes the first edition, first printing, of an original print-on-demand book?" For instance, what if the first order placed is for one copy of a given book. It is printed, bound, and shipped to its purchaser. Then, a week later a second copy is ordered. It, too, is printed, bound, and shipped to its purchaser. Does the first copy constitute the first edition, first printing, and the second copy the first edition, second printing? Obviously this question needs further study — I do not have the hubris to even try to propose an answer to this question. It is also obvious that print-on-demand books are going to change book collecting just as computers have changed manuscript collecting.